This is a decent excuse to shunt regulated the output stage. Use the big Ixys 100mA CCS set to 100mA, then let the other 30mA be your shunt reg operating current.
Actually, it could remove the risk altogether. Small value of series resistance that drops 24VDC, connect the coil of a relay across that and have the contacts switch 'something, somewhere else' that shuts off the output stage if/when there is no voltage drop across that resistor (no load=no drop).
Take a look at HY40. Not HY40Z, but the straight 40. Not too expensive either. If you find one and it is not good, let me know and I will give you a blank warranty card to send with it back to Hytron...
The HY40 is a power tube in its own right; its plate rating is probably higher than what is coming in 300B boxes these days.
cheers,
Douglas
The HY40 is a power tube in its own right; its plate rating is probably higher than what is coming in 300B boxes these days.
cheers,
Douglas
Last edited:
Not 300B in output, but my driver for GM-70 is:
3A5(parallel) - 4P1L (triode connected) - IT - GM-70.
All tubes with fixed bias. Working on CCS for 3A5.
I'm very satisfied with the sound. Measured on R&S UPL, driver stage is very linear.
I'm sure you can use this for info for your build.
Regards
3A5(parallel) - 4P1L (triode connected) - IT - GM-70.
All tubes with fixed bias. Working on CCS for 3A5.
I'm very satisfied with the sound. Measured on R&S UPL, driver stage is very linear.
I'm sure you can use this for info for your build.
Regards
Hello filkoman - changing the output tube is always an option as you say. I used to have a lot of 813 which I sold off - too much heat at 50VA and I didn't see the benefit over a 300b since I would have used it at low voltage anyway. And these big power tubes are expensive now. I'm actually very happy with my PSE 4P1L output stage, and the 300b is an experiment just because I have a bunch of them and it's a challenge!! But good lateral thinking here, and definitely a solution if you want a little more power for a bigger room.
100V 10mA. Bias about -2.1 to -2.3V. Depends upon tube. For now I have 180V B+ with 8K anode resistor. But will change to 10mA CCS only.
Sounds excellent.
Some years ago I played with same schematic.
The most important thing (I you use appropriate parts) is the driver tube anode choke.
If you use average, mediocre anode choke the result will be so mediocre.
Attachments
Interesting. Thanks for the heads up.
When I have money again, I was planning on trying these out for a 10Y preamp. Maybe they would make good plate chokes for the circuit in question? Shrug.
200H anode plate choke for 6sn7 6j5 5842 5687 12au7 -- One Pair | eBay
I'm careful (and very sceptic) with these far eastern things.
Without measuring (what someone independent DIYer made) I will only buy it if the price is marginal (buy it and throw it away if inadequate).
I keep buying -decent quality- alu chassis (from EU warehouse), but tubes or transformers ... that's another thing.
BTW 80H is "enough" for 10/801 (about 10k load at 20Hz), but interwinding capacitance is the most important thing.
P.s. C2 is reversed if it is polarised cap (I used polypropylene).
Without measuring (what someone independent DIYer made) I will only buy it if the price is marginal (buy it and throw it away if inadequate).
I keep buying -decent quality- alu chassis (from EU warehouse), but tubes or transformers ... that's another thing.
BTW 80H is "enough" for 10/801 (about 10k load at 20Hz), but interwinding capacitance is the most important thing.
P.s. C2 is reversed if it is polarised cap (I used polypropylene).
Last edited:
My idea was actually to buy two sets and run them in series like people do with the hammond 156Cs. This would double the inductance and halve the capacitance. Plus (if the DCR figures are correct) it would mean that the DCR would be juuuuuuust about right for biasing the 300b.
Sadly between a PM from audiowize and some subsequent research it does indeed seem like these valab chokes are iffy at best.
Hi Andy,
I am driving my 300B/842 currently with a 801A. IT is 1:1.
The DAC is a soekris and got as well a 801A, so input is 18Vpp...a bit more than what you got...which DAC do you use btw ?
I tried a LL7903...this is really a bad sounding device. Save your money.
I would instead ask our friend 50AE to wind an IT with 1:2 ( see here DHTRob - An AD1 preamplifier with Alexander Naydenov transformers ...I compare his 1:1 nano with monolith nano 1:1 and prefer 50AE.... and run them the 801 with 30-40mA for transient response.
Or use a 841 with the gyrator board...or maybe with a mu-stage ? The new version with reduced complexity or choke ontop of the 841 anode ? Or have a deep look into DHT which are typically overseen...there are some...
All the best
I am driving my 300B/842 currently with a 801A. IT is 1:1.
The DAC is a soekris and got as well a 801A, so input is 18Vpp...a bit more than what you got...which DAC do you use btw ?
I tried a LL7903...this is really a bad sounding device. Save your money.
I would instead ask our friend 50AE to wind an IT with 1:2 ( see here DHTRob - An AD1 preamplifier with Alexander Naydenov transformers ...I compare his 1:1 nano with monolith nano 1:1 and prefer 50AE.... and run them the 801 with 30-40mA for transient response.
Or use a 841 with the gyrator board...or maybe with a mu-stage ? The new version with reduced complexity or choke ontop of the 841 anode ? Or have a deep look into DHT which are typically overseen...there are some...
All the best
Last edited:
Exactly... he is as well on facebook I believe.
I forgot... rs242... t20.... 1623...all dht and a mu of 20...rs242 is oxid, the other tt. I tried eml 20...not sounding very nice in my system at least... 801a was much better... but it is important you get the heating right... coleman regs with 5v more from the raw supply than vdc target is.
I forgot... rs242... t20.... 1623...all dht and a mu of 20...rs242 is oxid, the other tt. I tried eml 20...not sounding very nice in my system at least... 801a was much better... but it is important you get the heating right... coleman regs with 5v more from the raw supply than vdc target is.
Good question - which DAC. I have a ES9023 right now. The application question could be solved in different ways in the DAC if it could provide more than 2v output. Possibly even some kind of solid state step-up device like a J-FET or whatever. My knowledge base is tubes, so solid state is a bit out of my comfort zone, but feel free to suggest.
"The ES9023 is a 24-bit stereo audio DAC with an integrated 2Vrms op-amp driver."...from their web page...op-amp...
So, I would try to get rid of the op-amp stage and use a DAC with a tube stage instead ;-)
So, we have 2Vrms, which is 5,8 Vpp, no ? We need 180Vpp for the 300B when it is biased at -90V. so a gain of 31.
To be honest, I would do it the same way i did it, which is universal tube output stage for the DAC and one stage infront of the end tube. Both 801. With this configuration you can play around with otherDACs and tubes as well, be it driver tubes or end triodes...use of IT vs. C-coupled stages etc. I will test next 814, 813,845, 211, 828, gm-70...all with the same topology, but with different voltages and appropriate OPt. I found that even within the 300B family there are big differences in resolution and tone.
Initially I have built the 801a dac output stage as a differential stage, so with four 801 and a cascaded current source, so that you connect the grids straight as well to current output dac like es9028, es9038. The grid to ground resistor does the i/v conversion. No C or IT in the signal path there...and than a LL1692AM as output transformer. doing PP to SE.
I converted to SE later as I find the distortion spectrum more natural and the Soekris is Voltage out anyhow.
A 841 stage with gyrator board or mu stage might work, but i would not forecast its microphony. Some 801A are mor microphonic than others and the 841 might be worse.
So, I would try to get rid of the op-amp stage and use a DAC with a tube stage instead ;-)
So, we have 2Vrms, which is 5,8 Vpp, no ? We need 180Vpp for the 300B when it is biased at -90V. so a gain of 31.
To be honest, I would do it the same way i did it, which is universal tube output stage for the DAC and one stage infront of the end tube. Both 801. With this configuration you can play around with otherDACs and tubes as well, be it driver tubes or end triodes...use of IT vs. C-coupled stages etc. I will test next 814, 813,845, 211, 828, gm-70...all with the same topology, but with different voltages and appropriate OPt. I found that even within the 300B family there are big differences in resolution and tone.
Initially I have built the 801a dac output stage as a differential stage, so with four 801 and a cascaded current source, so that you connect the grids straight as well to current output dac like es9028, es9038. The grid to ground resistor does the i/v conversion. No C or IT in the signal path there...and than a LL1692AM as output transformer. doing PP to SE.
I converted to SE later as I find the distortion spectrum more natural and the Soekris is Voltage out anyhow.
A 841 stage with gyrator board or mu stage might work, but i would not forecast its microphony. Some 801A are mor microphonic than others and the 841 might be worse.
Last edited:
In my 841 collection the ST ones (all Hytron, except one pair -maybe rebranded Hytron- RCA) are not tend to microphony (the worst ones are dead now with filament crushing), but globe pairs (Deforest, Taylor) are most sensitive to vibration.
These tubes age are 80-90 years, so socket damping and suspended sub chassis for sockets is mandatory.
BTW the 801 in the DAC stage and 801-300B is the three stage solution.
These tubes age are 80-90 years, so socket damping and suspended sub chassis for sockets is mandatory.
BTW the 801 in the DAC stage and 801-300B is the three stage solution.