The 26 with Ale's anode-load board should drive the splitter very well, if you take the output from the Source terminal.
It the 26 anode voltage is somewhere near the dc voltage that the splitter input sees now, dc-coupling should be good, too.
Put something like 470Ω at the output (if it has no source resistor already), to keep the FET from being tempted into oscillation.
Whether the gain is optimised - is best checked by doing it!
It the 26 anode voltage is somewhere near the dc voltage that the splitter input sees now, dc-coupling should be good, too.
Put something like 470Ω at the output (if it has no source resistor already), to keep the FET from being tempted into oscillation.
Whether the gain is optimised - is best checked by doing it!
Hi Rod,
Thanks for your reply!
When you refer to the Source Terminal, are you speaking of the connection labeled "anode" on Ale's PCB? This comes off the BSH111bk source (via a 470R). Just making sure I understand you correctly.
I did try the "Mu" connection, which is the junction of the BSH111 source and the gate of the XTP08N50D. But it was a bit strained sounding (a little squashed and static-y, no bass).
I also tried my LC plate-load version (I built two preamps) which I thought should be similar to using the "source" (Anode) out of the Gyrator boards. This sounded better, but the signal still became static-y sounding on anything but the quietest passages of music.
Would the Anode out of Ale's PCBs have more drive than a choke loaded 26? Should I give that a try?
It might be nice to get this to work. Kill a couple birds with a single stone. Though, I am contemplating a new amplifier build. Hmmm...
Thanks again for the time and knowledge!
Regards,
John
Euro 21- Thanks for the D3a loaded schematic. I filed that away for future contemplation.
Andy- If you have your IC OP amp socketed, and don't mind soldering an SMD chip to an adapter, you might try an OPA1641. I found it struck a nice compromise (e.g. between weight and detail) when I was playing around with chip-based RIAA sections. I only suggest this because I have never enjoyed the 4562 anywhere I have tried it.
Andy- If you have your IC OP amp socketed, and don't mind soldering an SMD chip to an adapter, you might try an OPA1641. I found it struck a nice compromise (e.g. between weight and detail) when I was playing around with chip-based RIAA sections. I only suggest this because I have never enjoyed the 4562 anywhere I have tried it.
Very kind of you to entertain my clumsy DIY attempts! This will no doubt reveal the absurd shallowness of my little puddle of electronic knowledge.
Attached is a quick cut'n'paste sketch of how I tried to hook things up. (it's a mish-mash of Ale Moglia's and Claus Byrith's schematics, altered to show how I have them built.)
Drawing this out makes me realize that I forgot the resistor to ground after the coupling cap in the one labeled Ver 2. Hmm.
Ver 1 didn't work at all. The music sounded thin, no bass at all, and the signal was broken-up and static-y.
Ver 2 was better, but still sounded like it was running out of steam. The signal broke up on any loudish or bass-y-ish passage. As when one cuts the power to the amplifier but let the music play as the caps discharge.
I have been toying with the idea of exchanging the ECC83 for a 6n6p/ECC99 Phase splitter which may work better as a 2 stage amp with the EL34s.
Maybe I should move this topic to another thread.
Last edited:
Hey John.
Sounds like you have some blocking distortion going on. The output from the “mu” is correct, that makes the 26 driver a mu-follower. The 26 should swing well the volts needed but perhaps won’t have enough gain in your system. It does sounds really nice so what you;re experiencing is probably something else in a design which you are experimenting on so could be many things.
I think you need to work backwards from the output section to check how much drive signal you need the 26 to provide.
Also the BSH111BK should have a gate stopper when you are running it below 10mA. Otherwise may oscillate in all sorts of different ways (LF hum noise, HF, etc.)
The original design you picked up on was meant to run under FB loop though.
If your original system had too much gain, perhaps you can experiment by adding an attenuator / volume control at the output of the 26 preamp before the original driver.
Cheers
Ale
Sounds like you have some blocking distortion going on. The output from the “mu” is correct, that makes the 26 driver a mu-follower. The 26 should swing well the volts needed but perhaps won’t have enough gain in your system. It does sounds really nice so what you;re experiencing is probably something else in a design which you are experimenting on so could be many things.
I think you need to work backwards from the output section to check how much drive signal you need the 26 to provide.
Also the BSH111BK should have a gate stopper when you are running it below 10mA. Otherwise may oscillate in all sorts of different ways (LF hum noise, HF, etc.)
The original design you picked up on was meant to run under FB loop though.
If your original system had too much gain, perhaps you can experiment by adding an attenuator / volume control at the output of the 26 preamp before the original driver.
Cheers
Ale
Hi Ale,
Thanks for your reply! The problems I am having are definitely due to some error in my implementation. I should first reiterate that it is my effort to bodge my amplifiers that is causing the trouble.
When I use your Hybrid Mu Followers and #26 feeding my amp normally (i.e. with the voltage amp in place) they sound really, really good.
With the NFB loop on my amps engaged, the noise I get is next to nothing. I just like the sound better without the negative feedback.
I was really just asking here to see if, in theory, running the 26+Mu follower straight into the phase-splitter should work. Then I could decide whether or not to keep trying.
If you think the BSH111 needs a gate stopper at less than 10mA, I will put some in. Good tip. I probably overlooked that in the documentation.
I should try it with my volume pot on the output. That would fix the bit of noise I get with the extra 20db of gain from not using the NFB loop on my amplifiers.
I, also, have an AVC, but it has never really convinced me. I know this runs contrary to established opinion, but I think resistive attenuators sound a little better, personally... for my tastes... probably just me...
Thanks again for your help.
Regards,
John
Thanks for your reply! The problems I am having are definitely due to some error in my implementation. I should first reiterate that it is my effort to bodge my amplifiers that is causing the trouble.
When I use your Hybrid Mu Followers and #26 feeding my amp normally (i.e. with the voltage amp in place) they sound really, really good.
With the NFB loop on my amps engaged, the noise I get is next to nothing. I just like the sound better without the negative feedback.
I was really just asking here to see if, in theory, running the 26+Mu follower straight into the phase-splitter should work. Then I could decide whether or not to keep trying.
If you think the BSH111 needs a gate stopper at less than 10mA, I will put some in. Good tip. I probably overlooked that in the documentation.
I should try it with my volume pot on the output. That would fix the bit of noise I get with the extra 20db of gain from not using the NFB loop on my amplifiers.
I, also, have an AVC, but it has never really convinced me. I know this runs contrary to established opinion, but I think resistive attenuators sound a little better, personally... for my tastes... probably just me...
Thanks again for your help.
Regards,
John
We can try to debug what's going on - I don't see why it should sound bad, at least up to the point of clipping (where blocking may cut in, as Ale mentioned).
The first step is to check and record the dc voltages at every anode and cathode, and at the mu output (but at the FET source). This last tells if direct dc coupling might be an option.
The first step is to check and record the dc voltages at every anode and cathode, and at the mu output (but at the FET source). This last tells if direct dc coupling might be an option.
Hi Ale,
That is extremely good advice!
Be methodical, work logically, and learn something as a result.
I have a tendency to bungle my way through things when it comes to electronics. A bad habit. Then when I run into problems and ask for advice, what I'm asking for, in essence, is free design work from others with some real knowledge. I don't have anything to offer in return, so this ends up to be very one-sided and selfish on my part (and very generous on the part of others!)
I feel rather badly about this. I think I shouldn't post here again until I hunker down and actually learn about what I am doing, and can contribute constructively to this forum.
In the meantime I have a very good HiFi to listen to (including two very good preamps: 26+Gyrator, and LC coupled 01A) and can, of course, enjoy this forum passively!
All the best,
John
Hi Rod,
Taking voltages on the amplifier side revealed to me that I cannot just pull a tube and expect it to work. (see the attached) This should have been obvious to me to begin with.
At this point the phase-splitters would need to be fully redesigned to do what I want.
I am going to do some reading (load lines, bias arrangements, etc.) and figure this one out. It will be good for me. I get my hands dirty all the time, but now I should concentrate on getting my brains dirty.
Best regards,
John
It works
So, just to update: the 26 with Ale's Hybrid Mu Follower has plenty of drive to skip the the Voltage Amp tube on a Mullard-style push-pull amplifier and feed the Phase-Splitter directly. (I'm sure this would work for the EL84 version, too). And it sounds very good.
The system is now nearly silent at idle, and I have much more reasonable range on the volume knob.
I don't think it would work correctly with the NFB loop engaged, but then one wouldn't really need to skip over the input tube if feedback was used.
I got the Phase-Splitter situation sorted out. Quite simple, really. All I needed was to create a voltage divider with the dropping resistor (just before the plate resistor) for the missing Voltage Amp tube so that the grids of the Phase-Splitter saw the same voltage as they would with the Voltage Amp in place. Now all the voltages look right.
I could have reworked the operating point for the 12AX7 to forgo having to supply any voltage to the grids (or, as Rod suggested, to work with the voltage provided by the Mu out and go DC coupled), but this was the quickest fix to implement, and the quickest one to undo if I decide I don't like it for some reason.
At any rate, it works just fine. I hate when problems, questions, etc. are left hanging in these threads.
Probably very few looking at the #26 pre amp thread would be interested, but I attached a schematic just in case.
Best regards,
John
So, just to update: the 26 with Ale's Hybrid Mu Follower has plenty of drive to skip the the Voltage Amp tube on a Mullard-style push-pull amplifier and feed the Phase-Splitter directly. (I'm sure this would work for the EL84 version, too). And it sounds very good.
The system is now nearly silent at idle, and I have much more reasonable range on the volume knob.
I don't think it would work correctly with the NFB loop engaged, but then one wouldn't really need to skip over the input tube if feedback was used.
I got the Phase-Splitter situation sorted out. Quite simple, really. All I needed was to create a voltage divider with the dropping resistor (just before the plate resistor) for the missing Voltage Amp tube so that the grids of the Phase-Splitter saw the same voltage as they would with the Voltage Amp in place. Now all the voltages look right.
I could have reworked the operating point for the 12AX7 to forgo having to supply any voltage to the grids (or, as Rod suggested, to work with the voltage provided by the Mu out and go DC coupled), but this was the quickest fix to implement, and the quickest one to undo if I decide I don't like it for some reason.
At any rate, it works just fine. I hate when problems, questions, etc. are left hanging in these threads.
Probably very few looking at the #26 pre amp thread would be interested, but I attached a schematic just in case.
Best regards,
John
How about removing the 0.47uf cap & 300K resistor, there's already a network on the Gyrator board or you can keep this & remove those from the Gyrator, worth a test & you should get better sound. Congrats on solving the problem.
Cheers
If it works nicely at 85V on the grid, the anode current in each triode is only about 0.5mA (seems quite low for a driver section). The current is set by the 82K resistor.
For fun, you could remove all those grid resistors (except for the stopper), and direct DC-couple to the gyrator FET's source.
precautions:
- up the 82K to 150K for the same current; optionally try 130K (0.6W MF resistors)
- HV supplies should come up about the same time!
For fun, you could remove all those grid resistors (except for the stopper), and direct DC-couple to the gyrator FET's source.
precautions:
- up the 82K to 150K for the same current; optionally try 130K (0.6W MF resistors)
- HV supplies should come up about the same time!
Hi Sumotan,
That would be better, for sure.
I did consider that, but I have other preamps I put in from time to time and most do not have a cap at the output. I thought it would give me more flexibility.
I should try removing the cap at the output of the 26/gyrator.
Good advice, however.
Best regards,
John
Hi Rod,
Yes, I have been mulling over your DC coupling suggestion (thanks for that!) while reading Norman Crowhurst—trying to get a better grasp on how to best approach this.
About the operating point: it actually sounds a bit better at ~70v. In Mr. Bryrith’s description he states that the anode of the voltage amp should be between 85-90 V. So this would be the grid voltage of the phase splitter, too, obviously. In practice the amp (with input tube in place) actually has 67v between the first two stages, and I think this sounds better.
This is what I am trying to understand right now with my reading. His section on DC coupling is a bit scant, so I will look at other literature.
Best regards,
John