I forgot about one DHT I have always wondered about but never tried. This is the STC 3A-108 that was a repeater tube. It's very linear, similar to 26 etc. in this respect. It exists in A version (American 4-pin bayonet) and B version (British 5-pin). Filament runs at 2V/0.25A. This has mu=30 and ra=50K in typical working conditions (130V anode voltage and -1.5V bias for about 1 mA) but it could be used up to 190V anode voltage. For the same 1 mA anode current bias would be close to -4V, or less bias for more current. It should have some flexibility for this application but it's not cheap....
andyjevans: "We haven't mentioned a concertina stage yet. Any fans of that?"
Concertina is the best in my mind next to transformer phase splitting. Concertina requires at least 3 stages in a power amp: input gain stage, concertina/split-load phase inverter stage, and output stage. But that would preclude for those who want only 2 stages in a PP amp.
Concertina is the best in my mind next to transformer phase splitting. Concertina requires at least 3 stages in a power amp: input gain stage, concertina/split-load phase inverter stage, and output stage. But that would preclude for those who want only 2 stages in a PP amp.
I've never heard that the 26 is designed to be non-microphonic. I stayed away from DHTs in general for years after reading complaints about hum and microphonics and the 26 was one of the tubes most often mentioned.
After I finally tried them (DHTs in general, not just the 26) I'm not sure what all the fuss was about. Maybe some people are just more picky than I am, though. As long as they don't exhibit any issues in normal use I'm good. I don't consider tapping on the tube to be normal use. And I don't mount them any differently than other tubes either. The most problematic tube I've tried has been the 12B4A, which is indirectly heated.
The 264 and 864 are relatively rare. I have some 864s and will be trying them in my breadboard project soon. I tried them when I had the Nuance breadboarded and they were nice but I preferred the 26, which are plentiful and inexpensive.
Yes, 3 stages. But the concertina doesn't amplify so it's not a gain stage.
Like a SUT is a "stage" and it does add gain but it's not a tube.
Anyone using a good concertina they can post?
I assume you mean to say "nearly impossible to make a 2 stage PP power amp with the Hammond + DHT driver". It works great in a SE amp but that's running it "backwards" as a 1:4 step up.
Running two in series (one as inverter the other as a step up) would obviously not work because you wouldn't get any gain. That's why I wondered if there was something that might combine the functions and have a similar step up ratio. I'm not sure a 1:2 would provide enough gain for typical DHTs that are mu 8 or less. Those Sowters also cost 3 or 4 times as much as the Hammonds.
But it seems that the 1140-LN-C could be combined with a 1:1 that can be configured as an inverter, correct?
45: Oh well. I thought the 1140-LN-C could be used as a step up. The goal is to use a low amplification DHT front end rather than any number of readily available IDHT tubes that fit the bill. Are there any Hammond products that would work in place of the 1140-LN-C? I'm from very close to where Hammond is headquartered, and Hammond is far more reasonable than other options,
Charlie: I have a pair of Jensen JT-11P-1 that could be combined with the 1140-LN-C, but a Hammond that does both would be ideal. I canceled my order for the 1140-LN-C for now.
As once suggested by Rayma in a post that in this instance a split-load circuit can have gain like below.
Explained by Merlin:
Hopefully others will comment on whether or not the Jensens are suitable or if there are other options. I'm sure there's more to it than simply being 1:1.
I'm curious how much the 1140-LN-Cs cost in Canada, where they are made. I bought one pair from a supplier in Chicago, which had the lowest price I could find in the US, and they were just over $100 each shipped. Then Andy suggested a dealer in England and I bought some more there. The crazy thing is that the cost, including shipping from England, was ~$45 each and they arrived in two days!
I found the data sheet for the JT-11P-1 and, although it is 1:1, I don't think it would work because it only has a single secondary winding. As I understand it you would need dual secondaries to use it for phase inversion.
Yes, I meant the amp we are discussing in this thread.
The EL34 Triode PP will need some 22-24V rms grid-to-grid depending on the actual conditions.
The 49 has mu= 4.5 at 140V/7mA. This is the actual mu of the tubes I bought a while ago (1.25mA/V with about 3.6K plate resistance). All curve traced! Because the available supply is 400V+ (necessary for the 15W class A the EL34 output stage) the anode load resistor resistor of the 49 can be 33K and still have some supply left for filtering/decoupling: 20V cathode bias+ 140V anode voltage + 231V anode resistor drop = 391V. That 33K is, say, in parallel with 220K next stage and will make the effective anode load 28.7K. So the effective gain of the 49 will be 4. With 3 V rms at the input of the Sowter you will get 6V grid-to-grid at the 49 and so 24V to drive the EL34. Input impedance 10K.
With the Hammond as phase splitter, 3V rms will become 0.75V at the secondary end-to-end and so you need an effective gain of 32 to get 24V drive.
The Sowter also has 5Hz-90KHz bandwidth. I am not sure the Hammond can do that but it looks good enough anyway.
Last edited:
That could be a solution, using a step up.As once suggested by Rayma in a post that in this instance a split-load circuit can have gain like below.
Last edited:
It could but it would become 2.2K input SE, in reverse mode. So one needs to add a concertina like @andyjevans and @directdriver suggest and to think more carefully about driving it as well. Let me have a look if there is something else.
Last edited:
With the SE input + concertina one could go for the 1140-LN-A which is 1:1.41 and use the 26 as differential voltage amplifier to get about the same sensitivity as the Sowter+49. And also the same 10K input impedance.
The only problem is that a NIB quad of 26 is not going to be as cheap as the 49. So the cost saving on the transformer will be diminished quite a lot....unless one already has the 26s.
I know because I only have one nice quad of 26's by RCA. All from the same batch and even the boxes are very well preserved and look like they were made not a long time ago....I have always thought about buying another quad but cost has increased a lot for really new (old stock) tubes.
The only problem is that a NIB quad of 26 is not going to be as cheap as the 49. So the cost saving on the transformer will be diminished quite a lot....unless one already has the 26s.
I know because I only have one nice quad of 26's by RCA. All from the same batch and even the boxes are very well preserved and look like they were made not a long time ago....I have always thought about buying another quad but cost has increased a lot for really new (old stock) tubes.
Last edited:
First Pass for Simple PP Amplifier
Since this may be the first time for some working with a PP amplifier the primary objective should be to keep it simple. That is what this one is, pared down to a minimum. That means substituting in tubes that are easy to work with in place of tubes of similar specs but without the warts. The amp in this case is fronted with 27s, their plate families are very similar to the 30. But there is no microphonic filament to create problems.
The 27s need 2.5 VAC of the heaters. This can be got from an unused 5VCT winding on the PT. If that is already used by a rectifier, then the 6.3VCT wdg will do. There are two ways to get the correct voltage. Refer to the attachments. The 27s are good for 90 VAC from heater to cathode.
The simulation uses the 6W6 triode connexion. This is close enough to triode connecting a number of TV sweep tubes to get an estimate of the amp performance. The Hammond 1140-LN-C is a step down transformer. The signal is reduced to ¼ of what it is at the input terminals. A 10V input results in 4 Watts of audio not including OPT losses.
The cathode bias is provided by common cathode resistors, not bypassed by an electrolytic. This allows for tube aging. And helps to assure the upper & lower halves of the cct have equal gain. For the keeners among you the connexion reduces even order harmonics. How it will sound is impossible to predict, just as with any other collection of parts. 😀
Since this may be the first time for some working with a PP amplifier the primary objective should be to keep it simple. That is what this one is, pared down to a minimum. That means substituting in tubes that are easy to work with in place of tubes of similar specs but without the warts. The amp in this case is fronted with 27s, their plate families are very similar to the 30. But there is no microphonic filament to create problems.
The 27s need 2.5 VAC of the heaters. This can be got from an unused 5VCT winding on the PT. If that is already used by a rectifier, then the 6.3VCT wdg will do. There are two ways to get the correct voltage. Refer to the attachments. The 27s are good for 90 VAC from heater to cathode.
The simulation uses the 6W6 triode connexion. This is close enough to triode connecting a number of TV sweep tubes to get an estimate of the amp performance. The Hammond 1140-LN-C is a step down transformer. The signal is reduced to ¼ of what it is at the input terminals. A 10V input results in 4 Watts of audio not including OPT losses.
The cathode bias is provided by common cathode resistors, not bypassed by an electrolytic. This allows for tube aging. And helps to assure the upper & lower halves of the cct have equal gain. For the keeners among you the connexion reduces even order harmonics. How it will sound is impossible to predict, just as with any other collection of parts. 😀
Attachments
Already discussed earlier in this thread, Hammond 124 Series Step Up Transformers.
Stuff it into the front end instead of the 1140, then we get plenty of signal, 👍
Stuff it into the front end instead of the 1140, then we get plenty of signal, 👍
directdriver,
If you like that modified concertina that has gain . . .
Just remember that the gain balance of the cathode circuit is not intrinsically the same as the gain at the plate circuit.
That is because the cathode circuit has to slew the capacitance of the floating interstage secondary.
That capacitance is effectively in parallel with the cathode resistor (the cathode signal resistor, not the self bias resistor and bypass cap).
With signal applied, the cathode has to move the whole secondary winding up and down.
At high frequencies, if you make the plate resistor and signal cathode resistor equal, there will be more gain in the plate signal, than there is in the cathode signal resistor.
The gain balance is dependent on frequency.
The beauty of the Original and True concertina is that the gains are intrinsically just as equal as the matched 0.1% resistors you might choose to use.
0.05% resistors, anybody?
I have owned amplifiers with concertina splitters, some of them worked good, some of them had Hum and Hash.
Then I designed an amp that used a concertina.
I no longer use a concertina in my amplifiers.
Just my opinions and experience.
If you like that modified concertina that has gain . . .
Just remember that the gain balance of the cathode circuit is not intrinsically the same as the gain at the plate circuit.
That is because the cathode circuit has to slew the capacitance of the floating interstage secondary.
That capacitance is effectively in parallel with the cathode resistor (the cathode signal resistor, not the self bias resistor and bypass cap).
With signal applied, the cathode has to move the whole secondary winding up and down.
At high frequencies, if you make the plate resistor and signal cathode resistor equal, there will be more gain in the plate signal, than there is in the cathode signal resistor.
The gain balance is dependent on frequency.
The beauty of the Original and True concertina is that the gains are intrinsically just as equal as the matched 0.1% resistors you might choose to use.
0.05% resistors, anybody?
I have owned amplifiers with concertina splitters, some of them worked good, some of them had Hum and Hash.
Then I designed an amp that used a concertina.
I no longer use a concertina in my amplifiers.
Just my opinions and experience.
Last edited:
I'm contemplating a PP build myself to use up a pair of PP OPTs. What I have in mind is double this SE design. I already built this in SE and it had plenty of gain and sounded very nice because of the DHT front end. The EL12n has a mu of 15 to 18 in triode. Available new and cheap from the big German electronics sites. No step-up needed. Very economical!
andyjevans,
Push Pull by doubling a single ended amplifier/
Needs a phase inverter to drive the two single ended inputs.
Then, one of the single ended amplifiers needs the primary winding's plate and B+ connections reversed.
Next, connect the Commons of the two secondaries together, and connect the two 8 Ohm taps together.
I did this once, and presented it a VSAC.
But you have to use 2 air gapped single ended output transformers.
You can only use the PP output transformers in push pull, not in single ended operation.
Using PP output transformers is for real push pull, or for Parafeed outputs.
Push Pull by doubling a single ended amplifier/
Needs a phase inverter to drive the two single ended inputs.
Then, one of the single ended amplifiers needs the primary winding's plate and B+ connections reversed.
Next, connect the Commons of the two secondaries together, and connect the two 8 Ohm taps together.
I did this once, and presented it a VSAC.
But you have to use 2 air gapped single ended output transformers.
You can only use the PP output transformers in push pull, not in single ended operation.
Using PP output transformers is for real push pull, or for Parafeed outputs.
I was using the term "double" loosely. I did mean a regular PP design with regular PP OPT. I was just too lazy to draw it all out. It would be balanced in with 2x 2P29L in filament bias. I'm just figuring out how to balance the input tubes. I can do this with the trimmer adjustment on the Rod Coleman regs if I use separate cathode resistors. I'd prefer to use one filament reg and cathode resistor if the 2P29Ls were already balanced. I do have quite a few of them. Trying to figure this out.
Alternative is a single 2P29L and a concertina.
Alternative is a single 2P29L and a concertina.