Original ones are ok, except the following ones:
R27, R28, R30, R31 = 100R
R29 = 150R
I recommend to change them in any case for increased stability margins.
R27, R28, R30, R31 = 100R
R29 = 150R
I recommend to change them in any case for increased stability margins.
I've noticed you use ceramic caps in most of your builds. Are they COG/NPO? I've been using mostly film or mica. Besides the fit issues, I'm wondering if this is where some sound quality differences are coming from with TubSoMo.
Those ceramic NP0 are either small value HF compensation or filtering, or HF decoupling. In any case, they influence the frequencies far above the audio frequency bandwidth, so... I don't really see (hear) the sound differences, caused by them.
The one at the input - 4.7uF or so - must be good. I use polypropilene film ones for that purpose. Non-polar aluminium electrolytics also work fine, but less reliable in the long run.
I use polypropylene for input as well. MKP if they will fit the board. The 4.7uF on the servo of the Vertical input was what was giving me the biggest fit issues, but I wasn't having much luck finding NPO to fit either.
I've got 11 TubSoMo input boards. I'll do a few builds with different parts and see if there's an audible difference. I should try some more tubes too, but I used the same ones you did for my first build.
I've got 11 TubSoMo input boards. I'll do a few builds with different parts and see if there's an audible difference. I should try some more tubes too, but I used the same ones you did for my first build.
I use the tubsumo ips/ops daily for some time now. Excellent amp. The new production tung sol 12au7 tubes I purchased for the build both became noisy (strong hum) after a month or so. I have a couple pair on NOS tubes. The first set I installed have been reliable for nearly a year.
someone bought up the idea of an aikido style front end with slewmaster style OPS...sounds interesting.
someone bought up the idea of an aikido style front end with slewmaster style OPS...sounds interesting.
I like this proiect !
It is beautifull .
Why not is a nonswitching bias solution , Hawksford , Sansui , Pioneer , JVC , Denon or Technics - old school type , ic LT 1166 ?
It is beautifull .
Why not is a nonswitching bias solution , Hawksford , Sansui , Pioneer , JVC , Denon or Technics - old school type , ic LT 1166 ?
I tried Tung-Sol and Electro-Harmonix in my first pair. Both sounded the same. It sounded like there was a low pass filter somewhere. High frequency response was badly attenuated for some reason.
My non-switching output stage is already in place - here is exactly the combination of TubSuMo front-end with NS-OPS:
TubSuMo IPS + NS-OPS test
Both options are excellent - the one with Lateral FETs is much simpler, the one with NS-OPS is even less distortion 😉 Although, with the LatFETs distortion is also very low (but it runs hotter for that).
Attachments
Hi Valery,
do I understand you to say that Non-switching amp runs cooler than the Latfet? Is that because of the large heatsink required to fit the board? The thing I like about the TubSuMo OPS is the smaller footprint.
do I understand you to say that Non-switching amp runs cooler than the Latfet? Is that because of the large heatsink required to fit the board? The thing I like about the TubSuMo OPS is the smaller footprint.
Lateral FETs show optimum performance at 120mA per pair quiescent current.
Non-switching runs great at 55-75mA per pair. That's the main reason - at lowish power, non-switching is barely warm, but lateral FETs are pretty hot (they like to run hot 🙂). Footprint - yes, LatFET OPS is rather compact. And simple 😎
As you can see, in the latest version I use different values for the gate stoppers in N and P shoulders, and don't use the gate-to-drain capacitors. Together with the changed R27-R31 in the IPS section, this setup gives the best performance and stability (comparing to the initial one).
OS and I both have noticed a temperature difference between the Sanken RET MT-200 temperature compared to a normal ON-Semi TO-247 as well. The Sankens run noticeably cooler at the same bias.
This is one of the simplest options possible.
Mine is temperature-compensated, with capability to adjust the clamping level for minimum distortion.
P.S. Lateral FETs are also great devices, when used properly. Expensive, but simple and great-sounding.
Wow it's really simple... is there a description somewhere about this layout?
Maybe I can fit this version of NS-OPS to my current layout... 🙂
Are these parts/values universal or they should be selected to the actual OPS?
I'm not sure if it is always true, but I have noticed with many of my amps that the ones with higher quiescent current run hotter at idle but once they are both driven, they tend to even out. I still need to finish the non-switching OPS so I can try it. Trying to convince myself to buy more Sankens. I still have a pretty big supply of the Onsemi MJL's.
I have the TubSuMo IPS's sitting in my lap so I can check for the values I have in it. It has been a while since I have listened to it' I'm looking forward to hooking it back up.
Hi Evan, great job!
I will check with regards to +/-85V and let you know.
Cheers,
Valery
On a side note, are you using MultiSim for all you simulation?
Excellent, Valery. Now that I learned Multisim I am ready to dig into Ultiboard 🙂 What are you using for PCB design? Do you have any netlists for vacuum tubes? Thanks.