A small friedman-inspired amp using an inexpensive power supply

[AX84CH]

I have been working on an experimental build that uses a 280v SMPS for power and a small 8k output transformer from a Randall RM20 for a pair of 6P3S tubes.

The preamp is Friedman-inspired and is the sound I am always chasing, not entirely unlike the AX84 Hi-Octane of old. 

Most of the parts were already on-hand and this was intended to be a low-dough build.  For the FX loop, I used a trick from Merlin's first preamp book for the send and a basic recovery stage.   

Global negative feedback is used to tame the overall tone and provide a depth circuit for some additional girth.  I didn't bother with a presence control since relaxing the NFB seems to do about the same thing for me on a lean amp like this.

It has turned out really well so far and has prompted me to start considering a version 2

Your comments and suggestions are welcome-

[passaloutre]

Care to share a link for your power supply? Also what is the capacitance multiplier?

[kagliostro]

@ passaloutre

https://www.electronics-notes.com/articles/analogue_circuits/transistor/capacitance-multiplier-circuit.php

---

@  AX84CH

Interesting build 👍

where did you get those PCB ?

Franco

[AX84CH]

Care to share a link for your power supply? Also what is the capacitance multiplier?

I think I bought mine from Ebay the first time but available from a lot of places.

https://www.amazon.com/dp/B0D9VCYBLB

[AX84CH]

@ passaloutre

https://www.electronics-notes.com/articles/analogue_circuits/transistor/capacitance-multiplier-circuit.php

Thanks for linking that.  Someone gave me a couple of prebuilt multipliers years ago.  I have read about them only at a high level and have used them in a couple of projects but have no real command of how to design them.

@  AX84CH

Interesting build 👍

where did you get those PCB ?

Franco

The helper PCBs and terminal strips were made at JLCPCB, though I have had OSHPark make them as well.   I really enjoy using the helper PCBs for projects because the let me have better heater layout than I would normally have and convenient point-to-point wiring around the tube socket.

Terminal strips are sort of expensive so I just started getting these little PCBs made and made a little closer to what I want in a terminal strip. 

All of this is a work in progress for hobby builds.

[kagliostro]

Thanks for the replay


Franco

[zircontweezer]

@AX84CH is that Chris Hurley? so much to thank you for if it is regards Joker/Joe.

[AX84CH]

@AX84CH is that Chris Hurley?

This is me. 

[JPK]

Welcome Chris! You got me started in building amps. I built the High octane kit back in 2010. The November shortly after. I'm up to 7 now. Interested in your thread.

[DummyLoad]

https://www.ebay.com/itm/267165428920

Seems similar to the unit you're using. No cage though.

--Pete

[DummyLoad]

Looks like this one:

https://www.amazon.com/dp/B0D9VCYBLB?tag=threadproductcarousel-20&linkCode=osi&th=1&psc=1&language=en_US

--Pete

[DummyLoad]

This is the one I'd zero in on... 30bux. 300V @ 300mA; enough for 20-25W amp without working it too hard.

Still, I prefer linear PS for tube amps. JMO.

https://www.amazon.com/Vacuum-Preamplifier-Switching-Transformer-100-265V/dp/B0DZGR1HPR/ref=pd_sbs_d_sccl_2_1/141-4689532-8312021?pd_rd_w=SY8Ej&content-id=amzn1.sym.aa738fbd-ad05-4d11-aae2-04b598db6305&pf_rd_p=aa738fbd-ad05-4d11-aae2-04b598db6305&pf_rd_r=PXC57AFWBQH84PVYSA2Y&pd_rd_wg=cdqpX&pd_rd_r=cf9d8605-54ab-40e5-9375-719858e5fb22&pd_rd_i=B0DZGR1HPR&psc=1

--Pete

[AX84CH]

This is the one I'd zero in on... 30bux. 300V @ 300mA; enough for 20-25W amp without working it too hard.

Still, I prefer linear PS for tube amps. JMO.

https://www.amazon.com/Vacuum-Preamplifier-Switching-Transformer-100-265V/dp/B0DZGR1HPR/ref=pd_sbs_d_sccl_2_1/141-4689532-8312021?pd_rd_w=SY8Ej&content-id=amzn1.sym.aa738fbd-ad05-4d11-aae2-04b598db6305&pf_rd_p=aa738fbd-ad05-4d11-aae2-04b598db6305&pf_rd_r=PXC57AFWBQH84PVYSA2Y&pd_rd_wg=cdqpX&pd_rd_r=cf9d8605-54ab-40e5-9375-719858e5fb22&pd_rd_i=B0DZGR1HPR&psc=1

--Pete

I have ordered one of these to try and should have it shortly.   I'm not sure it will be radically different though an extra 20v is welcome. 

I would probably prefer a linear supply as well, all else being equal.   If these end up being workable for a segment of projects, they are considerably less expensive and lighter than a linear supply.  Of course, everything is a tradeoff.  I'm pretty happy with the amp I built so I would say it has been a general success.

[DummyLoad]

This is the one I'd zero in on... 30bux. 300V @ 300mA; enough for 20-25W amp without working it too hard.

Still, I prefer linear PS for tube amps. JMO.

--Pete

I have ordered one of these to try and should have it shortly.   I'm not sure it will be radically different though an extra 20v is welcome. 

I would probably prefer a linear supply as well, all else being equal.   If these end up being workable for a segment of projects, they are considerably less expensive and lighter than a linear supply.  Of course, everything is a tradeoff.  I'm pretty happy with the amp I built so I would say it has been a general success.

I have one on my bench as I type this doing a burn-in. Have only a resistive HT load on it, 4x 300Ω 25W resistors in series; the stack measures 1190Ω total. With that load the PS delivers 285V, so the calculated current draw is ≈240mA - Meh! typical Asian over-reach spec hype. I just viewed a video another buyer posted just a bit ago and they were seeing similar results. So, probably will return it and leave a 2-3 star review with my eval.

Link to lab results

--Pete

[AX84CH]

I have one on my bench as I type this doing a burn-in. Have only a resistive HT load on it, 4x 300Ω 25W resistors in series; the stack measures 1190Ω total. With that load the PS delivers 285V, so the calculated current draw is ≈240mA - Meh! typical Asian over-reach spec hype. I just viewed a video another buyer posted just a bit ago and they were seeing similar results. So, probably will return it and leave a 2-3 star review with my eval.

Link to lab results

--Pete

Hey Pete-

I notice you don't have a load on the filament supply.   I don't know if it is the case on the one you have but the B+ on the one I have seems to behave better with a load on the filament supply.

I will get around to testing both of these a little more.

Also on the one I have, as I increase the current demand by biasing the amp hotter, the plate voltage is creeping up a bit.  I think it was 285 with ~110mA versus 280 with ~70mA

[AX84CH]

The one I got appears a little different from the one in your picture

https://www.ebay.com/itm/157252283200

[DummyLoad]

Yeah, I saw that same pattern PS on Ebay as well, seems like several accounts to same vendor, USA stock BS, etc., bulk ordered and shipped from China; why they state the 4-7 days ship transit time. 


On the unit I have, adding the 6.3V load did not change things. Test results in link.  3x 10Ω 5W resistors in parallel, for 3.3Ω total load, or about 1.92A current load. Some of the resistors were cooking at ≈130°C. 


--Pete

[johnnyreece]

Yeah, I saw that same pattern PS on Ebay as well, seems like several accounts to same vendor, USA stock BS, etc., bulk ordered and shipped from China; why they state the 4-7 days ship transit time. 


On the unit I have, adding the 6.3V load did not change things. Test results in link.  3x 10Ω 5W resistors in parallel, for 3.3Ω total load, or about 1.92A current load. Some of the resistors were cooking at ≈130°C. 


--Pete

It looks like your jumper is set at 280.  Thanks for sharing the testing results!

[DummyLoad]

Yeah, I saw that same pattern PS on Ebay as well, seems like several accounts to same vendor, USA stock BS, etc., bulk ordered and shipped from China; why they state the 4-7 days ship transit time. 


On the unit I have, adding the 6.3V load did not change things. Test results in link.  3x 10Ω 5W resistors in parallel, for 3.3Ω total load, or about 1.92A current load. Some of the resistors were cooking at ≈130°C. 


--Pete

It looks like your jumper is set at 280.  Thanks for sharing the testing results!

It was! Thanks for pointing that out. I was completely oblivious to the existence of that jumper   Seeing 304VDC now with the 1200Ω load bank.

Link to test results.

--Pete

[AX84CH]

This is the one I'd zero in on... 30bux. 300V @ 300mA; enough for 20-25W amp without working it too hard.

Still, I prefer linear PS for tube amps. JMO.

https://www.amazon.com/Vacuum-Preamplifier-Switching-Transformer-100-265V/dp/B0DZGR1HPR/ref=pd_sbs_d_sccl_2_1/141-4689532-8312021?pd_rd_w=SY8Ej&content-id=amzn1.sym.aa738fbd-ad05-4d11-aae2-04b598db6305&pf_rd_p=aa738fbd-ad05-4d11-aae2-04b598db6305&pf_rd_r=PXC57AFWBQH84PVYSA2Y&pd_rd_wg=cdqpX&pd_rd_r=cf9d8605-54ab-40e5-9375-719858e5fb22&pd_rd_i=B0DZGR1HPR&psc=1

--Pete

This time I have ordered one of those exact units to try.  The other one I got is considerably larger.  I could use the 300V without the doubling in size.

[passaloutre]

Yeah, I saw that same pattern PS on Ebay as well, seems like several accounts to same vendor, USA stock BS, etc., bulk ordered and shipped from China; why they state the 4-7 days ship transit time. 


On the unit I have, adding the 6.3V load did not change things. Test results in link.  3x 10Ω 5W resistors in parallel, for 3.3Ω total load, or about 1.92A current load. Some of the resistors were cooking at ≈130°C. 


--Pete

It looks like your jumper is set at 280.  Thanks for sharing the testing results!

It was! Thanks for pointing that out. I was completely oblivious to the existence of that jumper   Seeing 304VDC now with the 1200Ω load bank.

Link to test results.

--Pete

Living up to your username lol!

How clean is the output? Good enough for preamp work, or would it need some extra filtering?

[DummyLoad]

Well, the eBay unit I ordered tested with disappointing results. I strung up 6x 100Ω 100W resistors in series to simulate a ~500mA load, had to roll up the adj pot to get 296V output, but after about 30secs the PS shutdown. Not good for a device purportedly rated to 600mA. PS gets a little warm, maybe 5-10deg above ambient, so don't think heatsinking is the issue. Suspect that it would probably not shutdown with a lighter load. Messaged seller with results, let's see what they say...   

https://www.dropbox.com/scl/fi/w731plc5p5nkbivuqqfdp/eBay_300V_SMPS_test-IMG_20260507_031158965.jpg?rlkey=69oy6d9qjloqsfibcbpp5jpsk&st=26llxdrs&dl=0


--Pete

[printer2]

What I found in some of the SMPS's I picked up from China is that they do not deliver their rated output unless fed with a higher input voltage. A wallwart that is suppose to put out 2A puts out half when fed 120V but when I increased my variac to 160V it gave its rated current. If they are designed with 220V operation in mind they skimp on us 120V people.

[AX84CH]

I continue to have better luck than others it seems.

The first SMPS that I got continues to work, including when the amp is wide open into a reactive load for periods of playtime.   2x6L6GC seems to be the biggest that will power up from cold although the SMPS doesn't seem to like to power up again with both tubes installed from a short power off.  I can pull one and it starts back up and I can then pop the other back in.

It will not generally power up with EL34 under any circumstance unless I pull one or both first, let it start up and then plug them in.  I ran 2xEL34 (into 4k load) for about 20 minutes yesterday mostly wide open with no issues.   There is some noticeable heat from that which I should probably try to quantify.  It seems like I have one of those laser temperature things somewhere.  The SMPS isn't hot but it is warm.

The good news is that the amp sounds great despite the various compromises.

I have received a couple of inrush current limiters that I plan to install on the 6V output to see if the SMPS will start up with those blunting the cold current requirements of EL34. 

Mine is the 250/280v "transformer cattle" unit with no specs.   

[jbrrrrr]

I love seeing "output cattle bile machine" when I look at transformers on AliExpress

[DummyLoad]

I continue to have better luck than others it seems.

Are you Irish? 

The first SMPS that I got continues to work, including when the amp is wide open into a reactive load for periods of playtime.   2x6L6GC seems to be the biggest that will power up from cold although the SMPS doesn't seem to like to power up again with both tubes installed from a short power off.  I can pull one and it starts back up and I can then pop the other back in.

Interesting - The first PS is which model? Am assuming the Amazon/eBay unit with the full cage - Similar to this one?

I have received a couple of inrush current limiters that I plan to install on the 6V output to see if the SMPS will start up with those blunting the cold current requirements of EL34. 

Willing to share the brand/part number you plan to use?

Mine is the 250/280v "transformer cattle" unit with no specs.

Link?


Units similar to that eBay unit I bought looks like now are being sold on amazon for 3x the price - This one. I plan to test Printer2 reveal about needing 220-240V to meet spec. Have a 800W 120V/240V step up/down xformer on the way.   


If you purchased one of those eBay units with the 6.3-GND-12.6V LV option, maybe string the EL34 filaments in series to the 12.6V source? That should help limit inrush current to less than that of a single tube.

--Pete

[AX84CH]

I continue to have better luck than others it seems.

Are you Irish? 

Perhaps-  :)

This is the SMPS I'm currently using in 280v mode:  https://www.amazon.com/dp/B0D9VCYBLB/

The ICL that I used is:  sl10-2r005 (2ohms / 5A).  The SMPS will reliably power up with EL34 now and my heater voltage is about 5.9V.  That's a little lower than I wanted but it powers up reliably with all the tubes now so I'll take it.

I should have made notes on where the EL34's biased up.  I can't remember if it was 55mA or 65mA but it was the end of my bias range.  I played the amp cranked for about 15-20 minutes with no issues- it sounded great into a 1x12 with a Celestion Lynchback.

I've also been playing it with the various other tubes-  6P3S, 6P3S-E, JJ 6V6S,  Winged C 6L6GC.   I've got adapters for EL84's also :)

[AX84CH]

I have continued to reduce the cathode bias resistor trying to extend the bias range-  on the EL34 and EL84 I don't seem to have enough range to get rid of what sounds like crossover distortion. 

6V6S, 6L6GC, 6P3S/6P3S-E and even 6550 don't seem be as hard to bias out of that.  As I am searching for answers, it appears that sharp cutoff is a feature of the EL34 an may be part of what I'm experiencing.

The JJ 6V6S seems to be a really great fit for this chassis.   It sounds great even when it is being hammered and has an inexpensive heater budget compared to the other tubes.   The 6P3S-E might be my second favorite.   

[DummyLoad]

I have continued to reduce the cathode bias resistor trying to extend the bias range-  on the EL34 and EL84 I don't seem to have enough range to get rid of what sounds like crossover distortion. 

6V6S, 6L6GC, 6P3S/6P3S-E and even 6550 don't seem be as hard to bias out of that.  As I am searching for answers, it appears that sharp cutoff is a feature of the EL34 an may be part of what I'm experiencing.

The JJ 6V6S seems to be a really great fit for this chassis.   It sounds great even when it is being hammered and has an inexpensive heater budget compared to the other tubes.   The 6P3S-E might be my second favorite.

In Class A with 300V, EL34 want about 90mA at idle w/ g2 + cathode currents or 180mA for the pair. Spec sheet states 190Ω for cathode R under each EL34. Your PS may not have that kind of available power and may be current starving the output stage with the EL34's. In the datasheet attached, a 7KΩ load is optimum, 8KΩ will work with reduced power, don't recall if you mentioned what OT turns ratios are. Your current config is running with individual cathode resistors, and 2W 1K trim pots might be stressed; 2W pots have a max I limit, may be banging off that limit with EL34s? A 100-250Ω 5W WW part might work better. With the data sheet conditions met, it should make about 22W, slightly less, if you're measuring with a sine wave source.

--Pete

[AX84CH]

In Class A with 300V, EL34 want about 90mA at idle w/ g2 + cathode currents or 180mA for the pair. Spec sheet states 190Ω for cathode R under each EL34. Your PS may not have that kind of available power and may be current starving the output stage with the EL34's. In the datasheet attached, a 7KΩ load is optimum, 8KΩ will work with reduced power, don't recall if you mentioned what OT turns ratios are. Your current config is running with individual cathode resistors, and 2W 1K trim pots might be stressed; 2W pots have a max I limit, may be banging off that limit with EL34s? A 100-250Ω 5W WW part might work better. With the data sheet conditions met, it should make about 22W, slightly less, if you're measuring with a sine wave source.

--Pete

This project started with a tiny 8k transformer.  I have upgraded to a bigger 6k6/8k transformer and am currently running mismatched for 4k.

I was aiming for Class AB which was partly why I was using a 4k primary.  The loadline (according to an online loadline calc of unknown accuracy) for 8k (my other option) was pretty flat and looked like it would basically be impossible to put it into cutoff from there.  I have been pretty surprised by the tonal differences running the tubes into a 4k versus 8k load-  I shouldn't be, but this is one of the experience holes I have.

I'm concerned about the 2W pots-  I am not sure how much current it can manage when set to 0 ohms and I assume it isn't infinite.  The 5W wirewound pots I've seen are huge so I'm sort of pinched.   I have 10k and 20k/ pots for fixed bias next time around.

I'm pretty sure the SMPS can't deliver the current the loadline says the tubes will want in Class AB so the whole exercise is probably moot in terms of EL34.

The other tubes seem to sound and work fine so I probably just need to put this project to bed and either play the amp as is or move on to the next thing

[DummyLoad]

There's room in/on that chassis for a conventional linear supply PT, choke, etc., push the EL34s to 375-400V to get them up in class AB, call it wrap, or just live with the 6V6s. Personally, I'd just run with the 6V6 with the 8K OT and call it wrap - You stated it was pleasing with 6V6s, run with that?

I bought that eBay "250W" SMPS to run a pair of EL34 in P-P class A with MkII Boogie-like preamp; I have a leftover ClassicTone 7K Marshall clone OT and was wanting to use it for something other than another Marshall build. Looks like I'll adjusting the chassis size up.

On another SMPS note, I'm using a Amazon SMPS for a spin of a LND150 spring reverb unit, however, needs about 15-20H of inductance to get noise down to tolerable level, still, less than a $60 PT. I'm using the CE-Dist mini tanks.

https://www.amazon.com/Acxico-Voltage-Converter-Supply-Module/dp/B09ZV31X69

https://www.amplifiedparts.com/products/reverb-tank-mod-mn-8eb2c1b-medium-decay-3-spring-mini?srsltid=AfmBOoq1cK5A45VAzN4O20B4zkj7Z6SHWStvyuZ4dWzunfDf8x9l1hjOyVw

https://www.dropbox.com/scl/fi/b9fb77mj43et0ajx23pm1/IMG_20260517_130553815_HDR.jpg?rlkey=3covvquktbp5hrcmmnp1qqcdx&st=mva722y4&dl=0

https://www.dropbox.com/scl/fi/pukt7dkij27xgh0jnkro8/IMG_20260512_114058713_HDR.jpg?rlkey=vupv2ywanic4icyony2d6lp14&st=c2ou3yuj&dl=0
 
--Pete

[AX84CH]

There's room in/on that chassis for a conventional linear supply PT, choke, etc., push the EL34s to 375-400V to get them up in class AB, call it wrap, or just live with the 6V6s. Personally, I'd just run with the 6V6 with the 8K OT and call it wrap - You stated it was pleasing with 6V6s, run with that?

I bought that eBay "250W" SMPS to run a pair of EL34 in P-P class A with MkII Boogie-like preamp; I have a leftover ClassicTone 7K Marshall clone OT and was wanting to use it for something other than another Marshall build. Looks like I'll adjusting the chassis size up.

On another SMPS note, I'm using a Amazon SMPS for a spin of a LND150 spring reverb unit, however, needs about 15-20H of inductance to get noise down to tolerable level, still, less than a $60 PT. I'm using the CE-Dist mini tanks.

https://www.amazon.com/Acxico-Voltage-Converter-Supply-Module/dp/B09ZV31X69

https://www.amplifiedparts.com/products/reverb-tank-mod-mn-8eb2c1b-medium-decay-3-spring-mini?srsltid=AfmBOoq1cK5A45VAzN4O20B4zkj7Z6SHWStvyuZ4dWzunfDf8x9l1hjOyVw

https://www.dropbox.com/scl/fi/b9fb77mj43et0ajx23pm1/IMG_20260517_130553815_HDR.jpg?rlkey=3covvquktbp5hrcmmnp1qqcdx&st=mva722y4&dl=0

https://www.dropbox.com/scl/fi/pukt7dkij27xgh0jnkro8/IMG_20260512_114058713_HDR.jpg?rlkey=vupv2ywanic4icyony2d6lp14&st=c2ou3yuj&dl=0
 
--Pete

Oh yeah-  it's pleasing in a variety of ways.  I'm just in the weeds.   It turned out so good that I will absolutely build another one and have the SMPS sitting on the shelf ready to go.

I'd be interested to hear how that nixie tube power supply works for you.  I bought something similar to use for preamps and I never could get a tolerable noise level.  I could hear behind all the noise that the preamp was working but the noise was monstrous heh.   That was a different SMPS and a different project from yours entirely, of course.