Santa Claus DIY Tube Amp

viren schrieb:

Screamgigi ji, "Ustaads" are only in other peoples' eyes. Would love to meet you, since we have so many common interests. As you seem to travel quite a bit, how about a visit here to Delhi? Viren.

Viren-ji
Thanks for your kind offer to meet me. I do travel around but lately visits to rajdhani have become fewer. My friend Ansghu (diyadio)lives in Gurgaon and I am planning to meet him as well. He is probably one of the top speaker system desiger in the world. And brave too ! A pair of unsafe electrostats with 2KV flowing thro them was on "trials" for several years in his bedroom

I think I shall also take up your offer on the power transformer. I was hoping to use one of the redundant PTX with me. However, none of them are right for the job. Do I need to talk to Harish-ji directly or you can facilitate that?

Thanks again for you kind offers on both.

Screamgigi ji,

Firm up on the specs for the power transformer, and I'll get it made for you from Delta.

The size will be (LxWxH) 115x115x100mm.

Viren.

Here is a more complete schematic. PSU and CCS later. This time with the dreaded NFB We are making a variable NFB loop. It is not shown on here, because I can't find a 4 pole rotary switch in S-Spice from which this schematic is screen-grabbed.

Hi Screamgigi -ji,

Sorry I came in late, I saw the schematic, and referring to a resistor that connects from the plate of the first triode to the grids of the push pull splitter. The value is 1 M ohm. Is this correct? I mean the signal to the lower grid will be attenuated furthur since the ouput impedance of the first stage will be low compared to 1 M ohm. Pardon my lack of knowledge in valves, but I am just curious to know.

Keep up the good work and hope to listen to it some day.

Behram.

Viren-ji
Thanks. PM on way.

Behram-ji

Those valves perform the phase splitting action. There are several types of Phase Splitters. The simplest in called a Concertina, which is good for small pentode like EL84. But for the valves like 6L6 we need a much bigger voltage swing. We are using a Common Cathode type Phase Splitter. If designed carefully this combines the simplicity of a concertina and the ability to deliver large swings.

That IM Resistor combined with the 1uF capacitor fulfils an important function of the Common Cathode type phase splitter. Actually it’s a little complicated to explain and I am thinking of making an additional page on how the input stage and splitter work together. Especially in our case when they are directly coupled. So just a brief note here:

A +ve potential in the V1’s grid will lead to increased current and decreased plate voltage for the valve. BUT the voltage in the Common Cathodes (tied together) will increase because as you have pointed the voltage in the V2 grid is kept constant at the cut-off point of V1 grid. That 1M and 1uF network achieves this. Current in V2 will decrease because it grid get more –ve compared to the common Cathode. Eb of V2 will also increase with a +ve potential present here. If we are not using a CCS for this Common Cathode then we may like to have slightly mismatched plate resistors. In our case we have 33K and 37K respectively. There are reasons for this.

I have done a very confused job of explaining this . I know how its works but I guess to put it down requires a little more effort. I do realise this a job half done and we shall shortly have a discussion on types and functions of Phase Splitters.

We now move to our Power Supply Unit (PSU). Since Sivat-ji had instructed to put a description on PSU, we shall try meeting his demand.

Consistent with the simplicity of a tube amp, the PSU is a simple but critical module. They perhaps make the most impact on a power amp. So we should invest our best attention on the PSU, which is the engine that powers our amp.

So what is a good PSU for a tube amp? IMHO some of the perquisites are:

1.Excellent regulation; enough headroom.
2.Max attenuation of ripple and mainline noise typically less than 80mV for a PP amp.
3.Lowest resonance frequency
4.Lowest PSU impendence.
5.Cheap and simplicity; low parts count.

Since we are destined to run into the Valve vs Silicone rectifier debate, lets us just go further and discover why the above is not possible with tube rectifiers and why we reluctantly need Solid State.

Lets now recount what demands our amp is placing on the PSU. And remember folks, tolerance in valve tech plus/minus 20%. So don’t take the knives out if I mention 100 volt here and 120volts later : )


Output Tubes:

6L6-G x 4 qty.

We are biasing our output stage in Class A. As per the data sheet a modest Class A operating point is Plate Voltage (Eb) of 250DC and a quiescent current (Ik) of aprox 62mA per tube. This is our preferred operating point. In this condition we dissipate around 15 watts, which is well below the max of 19watts plate rating of the 6L6-G.

So our standing current draw for the output stage is 62mA x 4 = 248mA.

We also have the driver stage. Our two 6SN7’s will like around 6mA for each half.
Lets give them around 20mA. Our input stage 6SL7 wants aprox 2mA.

Thus our nominal requirement is approx 270mA. But we want to keep a few in reserve so lets beef it up by a factor of 20%. We are now looking at 270mA + 20% = 324 mA but 300 mA is more reasonable. That is a very comfortable and generous outlay for the main HV called the B+

But it is a little hard for a Tube rectifier

A popular tube rectifier like GZ34 has a safe current rating of 250mA. We can perhaps push that considering our low Eb. Howevere, other issues are more severe. Tube rectifiers have nominal resistance of typically 70~100 ohms. Some even have limiting resisters to protect them. When current start flowing thro a tube rectifier this series resistance not only waste electric watts by burning off the heat, it also makes an unpredictably poor voltage regulation. Tubes also have a limit on charging cycle, which mean a smaller capacitor. And the typical small value electrolytic caps don’t make good regulators nor do they provide enough cleaning. Exceptions are oil filled caps or modern MKP Motor Run caps.

On the other hand a SS diodes typically have a resistance of the less than 1 ohm and waste less on heat generation. They can usually take several times more current and HV and as result we can put in a larger capacitor leading to lower ripple, one of our original design goals.

Trust me, its hard for a tube guy like me to sing peens for the lowly SS but you can see why. There are other compelling reasons, often shrouded in tubelore, to use tube rectifier. But in this case we shall stick to SS for a combination of reasons; most of it being the simplicity and lower cost. But I cross my heart and swear with hand over flames that my 845 rebuild is seeing tube rectifier and just that is major project in itself.

We also wish to use a Choke in our PSU for better filtering. Chokes have a DCR and we need to take that into account when we aim for the final B+. Usually a good choke will have a very low DCR. Lets take a typically DELTA Choke model 2048. It has an Inductance (L) of 10 Henry and a respectable 75-ohm DCR. A tribute to DELTA’ quality. So we shall use this choke and call it the L1.

Now we need to find our load resistance (RL) that is connected to the PSU. As per our operating condition we need a main B+ of 250 VDC and a standing current draw of approx 250mA. As per Mr Ohm R=V/R which in our case is:

250v/0.3A = 833 ohm.

Voila! We thus have the most important factor that is required to make rest of our calculations!

We will now use ohms law to figure out how much AC we want from our Transformer. AC to DC conversion is typically 1.4 times the AC in voltage. Since our requirement is max 310 VDC before the Choke, let us factor the AC secondary voltage:

310/1.4 = 221VAC

So do we specify 220 volts as our secondary? Not so fast! Transformers have an efficiency rating. Typically 80% or so. If we specify 100 volts we may end up with only 80 volts under actually load. We shall be more careful and specify 275 volts AC as secondary at minimum 400mA. This will be rectified by silicon Ultra Fast diodes before being presented to our 1st capacitor which is C1. Our string will be like this

AC > BRIDGE > C1 > L1 > C2 > R1 > C3 > R2 > C4. This is called a Capacitor input PSU. We shall then find out, using the Ohm’s law, the value needed for dropping resistor R1 and R2 required to adjust the final voltage to all three stages on our amp. Easy isn’t it.

For all practical purposes we can start building a PSU at this stage. A simple amp like ours is not going to make any complex demands. But we want more exact numbers to see how our PSU meets the other design goals.

Our PSU’s regulation and filtering is highly depended on the specifications of the capacitor (C) and more specifically on its reactive resistance. The higher the C value the better the regulation and lower the PSU impendence. A rule of thumb is our C should have a reactance at the ripple frequency (ZC) of less than one tenth the RL. So lets see what is the situation in our case by using the formula = (RL)x(1/10)

In our case we must not cross a ZC of (833 x 0.1) 83 ohms max as per the rule of thumb stated earlier.

Now the absolute reactance of capacitance can be interpolated from this:

ZC = {(1000000)/(2pI x C x F)} when considering

ZC = reactance of capacitor
1000000 as the constant K
2pI (pye) = 6.28 approx
C= capacitance and
F= Lets say our supply freq of 50Hz

We can now solve that to find C.

C = {(1000000)/(2Pi x ZC x F)

Putting the numbers

C = {(1000000)/(6.28 x 83 x 50)

This works out to around 38uF and that is a fairly low C. Ripple voltage (VR) will be too high. This will be perhaps ok if we had no choice; however, we now have access to good elecos. It would be wise to put in more than 31uf.

But how much more?

Besides the technical reasons, there is a school of though that advocates only as much uF as required. Lower uF gives a faster and naturally airy sound, while the larger uF could mean a higher reserve of power translating into firmer bass response but may also cause a slower rise time. We will choose a happy middle ground. Lets assume we use 100uF. So what do we have now?

For a more correct figure (as per the above formula) the ZC of a 100uf cap turns out to be approx 16 Ohm. To verify this we revisit our “thumb rule” ratio of ZC:RL= 0.1 or less. In our case it is

ZC/RL
16 / 833 = 0.019

We are well below the golden ratio of 0.1 as dictated by wise ole men of yore. And don’t question that validity of that “golden ratio” I have no answer Old men have ruled that; lets not question them

So finally what is our VR with the chosen C? For this too we have a formula:

VR = (I x 4400)/ C where 4,400 is a constant for 50Hz mains and I is our current draw which is .3 amp.

Or

VR = (0.3 x 4400)/100 = 13.2V rms.

And the Ripple Current,

IR=VR/ZC
13.2/16 = 0.825 A rms

Note that this is directly proportional to C value e. i., if we have a larger C of say 220uf then the VR=6 vrms.

We have the VR and IR at the 1st capacitor. We now need to employ appropriate measures to clean that. Our chosen capacitor should enough Ripple current capability and should be more than IR, which is 0.825 A.

Out comes the dog-eared Keltron Passive Comp data book and we can quickly see that our preferred 100uf / 450 VDC capacitor has a rated ripple of 1.2 Amp. We are safe ! A word of caution here. Most tiny Chinese caps will have a much lower rated ripple current handling. That is surely asking for trouble.

What happens if we stick to our calculated C of 31uf instead of 100uf? Our VR and IR are up several factors. We will have trouble finding a 31uf capacitor, which is rated to handle 1 Amp rms. We can perhaps do that. AC motor run caps are rated very high. But our smoothening will be not high enough and we may need an additional pair of R and C. In the actuall build we shall experiment from 47uf to 470uf and find which value offer the most balanced sound.

Now we need to see if we are meeting our other design goal of max 100mV ripple and a low resonance freq.

Lets again see our figures:

Our C = 100uf
ZC = 16

Our VR = 13.2 vrms at 1st cap which is too high. We need to reduce it to about 100mV. We have a thumb of rule here too. It has been ordained that subsequent filter section will have an attenuation factor of at least 10. For example a VR of 10v rms at C1 should be down 1vrms at C2. In our case we need to get down from 13.2vrms at C1 to around 100mV at C2 where it be fed to the amp. That’s a tall order and we shall see that it is done elegantly.

To really do it well we shall put a Choke beween C1 and C2. we can find the value of this choke easily done by finding the absolute reactance of our caps, which we shall call ZL, dividing it by our required attenuation factor and then using another constant K (not me ! blame the old men) which shall give us the number of Henry required. In our case it should be around 2 H but we have already decided to use 10H cause it is available with us and it should be enough.

There is another benefit of using a larger H. This helps in keeping our frequency of resonance low. We ideally like it around 5Hz and well below our lowest corner frequency. Just to soothe our nerves we shall do a quick check if a 10H Choke is meeting that requirement.

Since our C is 100uf and we want a Freq of Resonance of less than 10Hz (lets assume we want 5Hz). Then we have this formula.

5035 Squared/Fr squared x C where 5035 is a again a K

What do we have in our case?

(5035 x 5035)/((5 x 5) x 100)) = 10140 mili Henry which is 10H. Bingo ! In practical world we will be happy if our resonance fr is around 7~8Hz and our calculations say that even 2H is enough to reach that goal

Actually we have several more factors to consider. However for all our practical purposes we have designed our PSU. Any more explanations will only make it dense and provide an opportunity to Amp_Nut-ji to declare, “oh tubes are so complicated “

For this post we have drawing heavily from that tube technology bible called Radiotron Designers Handbook (4th edition) published by RCA in 1952. It’s got more than 1,500 pages of dense but interesting design tricks for tube gear. Just the TOC runs 50 pages PSU and related things are covered in Chapter 30. Worthwhile read.

I expect a whole bunch of mistakes in this post. I did not have time to review and will feel grateful if obvious mistakes are pointed out. Next we shall post the schematic and explain the various windings.

And here is our PSU schematic draft.



First the Transformer load:

HT: 275 volts x .4mA = 110 watts
LT1 : 6.3 volts x 6 A = 37 watts
LT2 : 50 volts x 0.05A = 2.5 watts

Total Secondary Load = 150 watts
Add 10% transformer loss = 15 watts
Total Max Load = 165 watts

This will allow us the calculate the mains fuse required by using the formula

Total Load/Mains supply voltage = 165/220 = .8 Amps

We will add 50% margin to avoid startup blowouts and feel that a fuse rated at 1.5A should be all right. But remember a fuse is expected to behave nicely. It will do its best to protect itself so that our other equipments down the road burn out fast

We have not calculated the values of dropping resistors. Once we get our power transformer we shall measure its impendence. That is required to make a correct prediction of our voltage readings.

screamgigi schrieb:

Besides the technical reasons, there is a school of though that advocates only as much uF as required. Lower uF gives a faster and naturally airy sound, while the larger uF could mean a higher reserve of power translating into firmer bass response but may also cause a slower rise time. We will choose a happy middle ground. Lets assume we use 100uF. So what do we have now?

Screamgigi,

your treatise on the PSU is indeed remarkably good. Thanks very much for the same. I did about 75% of it but I need to go back & re-read it & digest it better.

Re. the supply ripple caps - one trick that many DIYers play is that they use bigger caps for filtering the ripple but ALSO bypass this big capacitor with a smaller capacitor typically 1/10th the big cap's value. This bypass cap is made of some high quality film (no electrolytic). This smaller cap provides those fast transients while the bigger caps provide the charge reservoir.
Might I suggest this technique for your PSU as well?

bombaywalla schrieb:

.... bypass this big capacitor with a smaller capacitor typically 1/10th the big cap's value. This bypass cap is made of some high quality film (no electrolytic). This smaller cap provides those fast transients while the bigger caps provide the charge reservoir. Might I suggest this technique for your PSU as well?

Yes I do that at times. It as its sort of a tweak. I usually parallel 1uf and .1uf MKP with the big PSU caps. I can see the difference in the scope. But this is a hands on kind of thing and I was planning to mention it at the time of actual build.

The Bypass capacitor thingy has been debated ... sometimes to the extreme.

There is one poster who INSISTS that the bypass MUST be with 1% of the original value ONLY, with atleast 2 levels, preferably 3 levels of bypass for best results.

Amp_Nut schrieb:

The Bypass capacitor thingy has been debated ... sometimes to the extreme. There is one poster who INSISTS that the bypass MUST be with 1% of the original value ONLY, with atleast 2 levels, preferably 3 levels of bypass for best results.

Yes I have seen those too.

The only gripe I have is the 1% thing. A bit point less considering an avarage big size electolytic capacitors will have a published tolerance of +/- 20% but often more likely to be around 50% in actual.

With Oil filled capacitors the bypass make no difference. I dont know why.

Best is to install and tweak the end result.

The bypass capacitor debate also extends to not having the bypass at all. Some valve designers argue that small value film caps sometimes form a resonance tank circuit with big electrolytics, actually adding a glare to the sound.

All I can say is, I have tried it both ways. I prefer the sound without any bypass caps at all!

Viren.

I found this site that provides a Great primer on Tubes.

Recommended reading for those who want a basic understanding of tubes.

Sorry folks I was a little tied up in my day job activities. But the good news is that we now have our base ready. It is with me and this Sunday I will drill out the holes and installed valve bases and the tag strip. Once its is done we shall start the wiring

I have also PM'd the Power Transformer specs to Viren-Ji. So now concrete action taking place.

On a more sober note, Shri GJO (of diyadio) is coming down from bangloroo to see me on Friday. He is the grand Tube messsa of Indian DIY scene and I just hope after that 2500km flight, he is sufficiently humoured on this amp. He has been following it rather keenly. Dont know if he posts here or not In any case a round of golfing is lined up, if the amp is too disappointing for him

Amp_Nut schrieb:

I found this site that provides a Great primer on Tubes. Recommended reading for those who want a basic understanding of tubes.

Which site? I think you forgot to provide the link.

The Classic RCA Raditron Manual Version 10 from the 1930's,
can be downloaded as an 8 MB pdf file from

http://www.nmr.mgh.harvard.edu/~reese/RC10/rc10.pdf

Amp_Nut schrieb:

The Classic RCA Raditron Manual Version 10 from the 1930's, can be downloaded as an 8 MB pdf file from http://www.nmr.mgh.harvard.edu/~reese/RC10/rc10.pdf

Actually the 4th edition is more applicable. I have a 77MB scan which is very neatly TOC's and linked. I will be glad to send a CDROM to any one interested.

As for me I have the real thing. Bought for Rs. 50 in Free School Street. It beats the CDROM. You can't take the laptop to the loo

As for me I have the real thing. Bought for Rs. 50 in Free School Street. It beats the CDROM. You can't take the laptop to the loo

Screamgigi....in todays day and age...I would'nt bet on it !
Some tech geek may have fixed a laptop with wi-fi in the loo with a wall mounted stand

Manek.

Amp_Nut wrote: I found this site that provides a Great primer on Tubes. Recommended reading for those who want a basic understanding of tubes. Which site? I think you forgot to provide the link.

Zillion Apologies, screamgigi.

The link is :

http://www.bonavolta.ch/hobby/en/audio/t_bas.htm

However, as our forum Guru on Valves, I am sure its 'Old Hat' for you.

Actually the 4th edition is more applicable. I have a 77MB scan which is very neatly TOC's and linked. I will be glad to send a CDROM to any one interested. As for me I have the real thing. Bought for Rs. 50 in Free School Street. It beats the CDROM.

I SO agree with you on the convenience of a Hard Copy.

That is why I pay so much more to read magazines in print rather than the electronic versions.

Let me check out the RCA manual that I have downloaded. If it floats my boat, I will take you up on your offer for the later version on a CD.

Thanks !

Screamgigi-ji,

What are your thoughts on a Regulated Power supply for Valve Pre and Power Amps ?

Do you feel they improve the sound ?

There seem to be varying opinions on this.

Also, Siva & Viren, your thoughts on this ... Much appreciated.

Thanks

Wow ! Fantastic stuff going on here

I am a newbie to valves and most of it goes over my head….

BTW: What is a valve rectified power supply?

screamgigi schrieb:

You can't take the laptop to the loo :P

aah , but you can do that with a Blackberry gets enough emails of the way during a most unproductive otherwise time

nuff said

Amp_Nut schrieb:

Screamgigi-ji, What are your thoughts on a Regulated Power supply for Valve Pre and Power Amps ? Do you feel they improve the sound ? There seem to be varying opinions on this. Also, Siva & Viren, your thoughts on this ... Much appreciated. Thanks

Amp_Nut, I'm also interested in what these folks have to say on this matter.
So far, I've not seen any regulated power supply for a tube power amp (& even tube pre, for that matter).

When I read/hear the word regulated power supply then the phrase negative feedback comes to mind simultaneously! that's one of the best ways I know to create a regulated power supply. However, having feedback wrapped around it has its cons when it comes to audio. When the power supply is regulating, its bandwidth has to be limited for effective regulation. Also, performance of the active circuit sets the performance of the down-stream gear. WIth unregulated power supplies, the components are passive & their performance/distortion is less of an issue. One has to left to deal with simply noise & any parasitics. This, IMHO, is an easier job than dealing with an active power supply.

One could fudge the situation & get a, say, PS Audio P300 or P1200 power plant to regulate the wall AC supply & then feed that to the pre &/or power & one would have a pseudo regulated power supply!
I've heard the P300 unit & it does well for certain front-end components but more often than not I've seen amps connected directly to the wall esp. if they are large tube amps.

Anyway, let's see what the people-in-the-know have to say about this..........

Amp_Nut schrieb:

Screamgigi-ji, What are your thoughts on a Regulated Power supply for Valve Pre and Power Amps ? Do you feel they improve the sound ? There seem to be varying opinions on this. Also, Siva & Viren, your thoughts on this ... Much appreciated. Thanks

Theoritically Regulated Power Supply would be good for all audio circuitry. However, the problem is - today's limitation of the regulation circuitry & components available!!

There are portions of the amplifier circuitry that might demand sudden surge in current flow..based on variation in music. Most regulation circuitries will NOT be able to keep up with such surge and drop in current, hence - a negative impact on the sound. The second problem is head room provided by regulated circuitry compared to unregulated power supply....this becomes a cricital factor (and a practicallity issue) especially in power amplifiers.

I think there are quite a few preamps (including valve based) that uses regualted power supplies. This should not be a problem in my opinion. I'm not aware of which brand uses regulation...but a lot of us DIY'ers do use them.

Regards
Siva.

Thanks Siva,

What you have penned is a good summary of the Regulated vs Un-regulated debate.

Can some others, address the issue, PARTICULARLY related to Valve Pre & Power Amps ?

As an example, let me point out :

The BlueBerry Valve Integrated Amp rave reviewed in the Oct Stereophile, has the designer Insisting that an Unregulated Power supply is essential for its natural sound.

The VTL 400 Valve Power Amp that is Pretty much State-of-the-art insists that the KEY contributer to their performance is the VERY tight regulation on their Power supply....

A DIY pal of mine who has made an OTL tube Power amp, says the sound is MUCH improved with a B+ PSU that has been regulated to a few tens of milli-Volts !


I know that there will never be a difinitive conclusion on this matter, but would love to hear from tube Gurus like screamgigi-ji & Viren.

Amp_Nut,

I'm no valve guru...but based on my limited experience, here are my observation.

What VTL does is one notch above what most of us conceive as regulation - an advanced auto-biasing circuitry (possibly micro-controller based). I would leave VTL and similar product out of this discussion.

Taking into consideration only - "simple" transistor based regulation - I think that it would not be wise to use them for B+ supply in a power amp. Regulation might sound good in some combination, but i would assume that when the power amp is subject to a difficult load and playing low freq....you are bound to face issues. I will agree that unregulated power supply (in power amp) is a better option....unless i can afford VTL.

Initially when i made my preamp, i thought 7812 would the suffice the requirement of the 12AX7 heater circuitry. When i was forced by my uncle to try the LM317, i was unable to believe the difference in sound. I think the head-room is a very important factor, that in turn would affect other aspects such as the impedance. I do use a mosfet based regulated B+ for my preamp, which is better than the unregulated one.

Eager to hear from SG (its very difficult to type "ScreamGigi" ) and Viren

Regards
Siva.

Hi Sivat,

Thanks for sharing.

1. The info that the LM317 made a Significant diff on the HEATER supply is a real eye opener. I will file away that info in the corner of my brain, forever ..

2. A regulated supply, is believed by some, to praticularly Improve bass, by tightening it. Tests usually show a lower distortion, throughout the audio band, but Particularly in the bass, when the supply is regulated... atleast in Solid State equipment, that I am familiar with. Subjectively, the bass is better defined, and better damped.

3. The VTL 400 Valve Power Amp does use 'trick' bias, but it Does use a regulated Supplu for the B= supply.

There is a lovely review in Stereophile at :

http://stereophile.com/tubepoweramps/1205vtl/

From the review, I quote VTL's Luke Manley:

"But the biggest thing is the precision regulated power supply. Along with maintaining correct bias, the way to keep tubes at their optimum operating point is by regulating the high tension power supply. Normally, the voltage begins to sag as power is pulled out of the supply, and B+ voltage might drop by as much as 10%. The S-400's supply never varies by more than 0.1% or 0.2%. "The S-400's supply is essentially a power amp in itself. The regulation requires six 3300µF capacitors and five 800V, 30A MOSFETs. They've got tremendous transconductance, so when the caps dump, the regulation keeps the voltage propped up. The screen grids are particularly sensitive, so they're actually double-regulated. Their supply is taken off the regulated plate supply, stepped down, and further regulated at the lower level."

Incidentally, for those not familiar with VTL S-400 Reference power amplifier, it has a Retail price of US $ 20,000 and the S-400 can be run in tetrode (300Wpc/8 ohms) or triode (150Wpc/8 ohms) modes.

It is a Full Valve amplifier ( NOT a Hybrid ) with MOSFETS performing the B+ regulation.

Would LOVE to hear it .... someday...

Amp_Nut schrieb:

"The S-400's supply is essentially a power amp in itself. The regulation requires six 3300µF capacitors and five 800V, 30A MOSFETs.

Wow !!

Screamgigi, Viren, please design a regulated PSU with this kind of capacity ...for our use.

Regards
Siva.

As usual in tube hocus-pocus, we have diverse opinion.

Unregulated supplies:
-Simplicity: Transformers, rectifier and filters.
-You can’t go wrong unless you try really hard.

Unregulated supplies:
-Complicated. A regulated PSU will have additional devices to sense the variation in set voltage and actual voltage being delivered. These devices will then “regulate” their conductance so that the delivered voltage matches the set voltage.
-You can really go wrong if you are not careful.

Why regulate?

Voltage stabilization becomes essential in select applications because the mains voltage is not always constant. It can vary 20% depending upon the load and time. Such variation will affect the performance and reliability of the tubes and change all measured parameters. It may even lead to tube failure if max op-point are exceeded.

I will most certainly NOT use regulated PSU in the out put stage of Class A or even Class AB amp. It will defeat the purpose and sound muddled

But situations where I may use a regulated PSU are:

-Screen supplies of a Pentode or Tetrode output stage
-Input stage of a power amp, particularly when I wish to DC the stages.
-Line stage, Tube Crossover
-And most certainly in a RIAA Phono stage.

All the above are critical application and do well if fed with a regulated, low impendence PSU. For instance a 10% change in voltage for a Phono stage will reflect in a 3dB variation on the RIAA standard. In other areas unregulated supply on a fluctuating load will overlay its own variation on the audio signal. Tubes essentially modulate the steady DC power supply with the varying AC audio signal. Unregulated DC supply will interpolate with the musical signal and result in higher distortion and most certainly weak bass response.

It’s to easy to screw up an Unregulated PSU to the extent of damage the equipment. Design, construction, placement of component, layouting all play an important rile in a regulator. Its easy to bring in the regulator to the audio signal itself. So a lot of care is needed.

My preferred method for an unregulated PSU will have a tube rectifier, choke and a gas tube shunt regulator. I surely will keep away from Zener and MOSFETS. They can induce a lot of artifact on the main rail. Besides I love that psychedelic purple glow of a gas tube

As Sivat-ji rightly say, regulated PSU is good if implemented well. As with other parts of a tube amp, there is no absolutely no sacrosanct method of building a PSU. All designs are at best a compromise. Key is how you compromise and to what extent.

Amp_Nut schrieb:

five 800V, 30A MOSFETs.

Do we know what these could be ... or somthing with this kind of capacity

sivat schrieb:

Amp_Nut schrieb: "The S-400's supply is essentially a power amp in itself. The regulation requires six 3300µF capacitors and five 800V, 30A MOSFETs. Wow !! Screamgigi, Viren, please design a regulated PSU with this kind of capacity ...for our use. Regards Siva.

And electrocute the entire neighbourhood

Hi Sivat,

I did not say that !

VTL's Luke Manley said that !

I agree the specs of the MOSFETS are ... awesome !
Often manufacturers get their own Semi conductors made. Sometimes these are a sort from the production like.

I did read an interview of KRELLs CEO who said that they had the KRELL output transistors designed & made from scratch.... the order quantity they confirmed : 1 Million pieces....

No wonder they make HUGE amps with LOTS of output decives.... they have to move them FAST !

screamgigi

My preferred method for an unregulated PSU will have a tube rectifier, choke and a gas tube shunt regulator.

Thanks SG for a nice post.

I am new to tubes...

Are Gas Regulators still available ? What kind of price ?

A Few Part Numbers and their key parameters ?

Any web links on the topic too would be nice.

Thanks

Amp_Nut schrieb:

screamgigi Are Gas Regulators still available ? What kind of price ? Thanks

Amp Nut,

I have some RCA OA2...i think i paid around $12 each. I did try them with my preamp....but preferred the mosfet regulation and solid-state rectification (sorry SG...personal preference). A friend of mine actually swears by the version on my preamp that i build using this gas regulation...which seemed to go very well with his Rotel.

I think you also get in locally in the market...

About the earlier post on VTL. I truly believe that VTL has built such an PSU...so please do not think my message was Sarcastic. If VTL can do it....i think we should also attempt it. First we should find out... what mosfets are available with such capacity...SG, Viren any ideas.

Regards
Siva.

Amp_Nut schrieb:

Are Gas Regulators still available ? What kind of price ? A Few Part Numbers and their key parameters ? Any web links on the topic too would be nice. Thanks

In India the cost of gas regulator tubes ranges from Rs. 25 to Rs. 75. Unlike other tubes, there is no demand for them here anymore. BEL made a lot of them. I use both BEL and RCA. They are easier to implement and look cool. Some say they are better, so we shall leave it at that

As for parts number and parameter, wikipedia is your freind Here, short and simple.
http://en.wikipedia.org/wiki/Voltage_regulator_tube

Amp_Nut schrieb:

"The S-400's supply is essentially a power amp in itself. The regulation requires six 3300µF capacitors and five 800V, 30A MOSFETs.

I have a strong feeling that this might be a Motorola power device. Motorola is "notorious" for making such honking big devices that end up in the audio industry. I have seen this trend before. Check the On Semiconductor on-line catalog.

There might be other companies that also make such devices - the same companies that make soft-recovery, ultra-fast FREDs such as IXYS, General Instrument, Harris, Fairchild to name a few. For example the FREDs that I have in my s.s. amp bridge-rectifier PSU are IXYS devices rated at 1200V, 60A. I'm almost sure that IXYS will also have power MOSFETs devices as the ones cited above. FWIW.

i hve been hit by a major med issue. i had a mishap and damaged my right palm small fingure. i think the actual construction of this amp will be delyed. sorry. i will look up the forum but may not be able to type much for next 15 days atleast. i need 7 days to build the amp.

screamgigi,

sorry to hear aboput the incident. Take your time GURU. No rush, take care of your hand and take your medicine on time

manek.

screamgigi schrieb:

i hve been hit by a major med issue. i had a mishap and damaged my right palm small fingure. i think the actual construction of this amp will be delyed. sorry. i will look up the forum but may not be able to type much for next 15 days atleast. i need 7 days to build the amp.

Screamgigi,

Take care...and get well soon.

Regards
Siva.

Guruji...

Very sorry to hear that....

Aaram Kijiye !

This is a Hobby, that should be savoured & enjoyed.

Relax and give it all a break.

Above all, wishing you a speedy and full recovery.

Take Care

very sorry to hear that..there are many things more important than music, and health is definitely one of them.

please do take care and rest so you can get back to us ASAP

wish you a speedy recovery, Screamgigi.

Screamgigi ji,

Sorry to hear about your mishap. Hope you recover soon.

Viren.

Screamgigiji,

Hope this does not mean that you will not be able to indulge in wrist exercises of this variety -

Here's wishing you a speedy recovery.




Krish

Dhanyawad bandhu-gan.

goodnews- i can type with left hand index finger
badnews- this hard plaster will be in place for 15 days more. that puts end to a plan to do some heavy stuff like drilling and filling the chassis.

i seriously think renaming the thread as halloween amp.

krish-ji; where there is a will there is a way. cheers. i assure u that those exercises are being performd diligently

Get well soon, Screamgigi.
With the addl time ,you may well be able to think up some breathtaking mod to the amp.
All the best and do keep up.
Kamal

well after all the christmas presents have been opened and time has passed and we're dreading getting back to work, in comes the amp! sort of extending the christmas spirit... so in a way it'll be perfect timing.

meanwhile get well lille finger

screamgigi schrieb:

i hve been hit by a major med issue. i had a mishap and damaged my right palm small fingure. i think the actual construction of this amp will be delyed. sorry. i will look up the forum but may not be able to type much for next 15 days atleast. i need 7 days to build the amp.

Use the opportunity (yes, misfortune is also an opportunity) to listen to some great music. It will be nice to be on the other side and listen instead of building. Maybe one of the Bombay forum members will be enterprising enough to help you build your amp. I am not in Bombay, else I would have taken dictation and typed for you. I love your detailed posts, which I regularly read. I do not post myself as I have insufficient knowledge on valves.

how are your now ?

hope the wrist exercises have been good for you