Dear all,
I want to do some mod to improve the sound quality of my mono-block amp, as attached.
As suggested, a change of high quality of OPT would be an option.
As shown, it is a PP with parallel EL34 for each path.
Lundahl LL1663 with 5k:8ohm would be possibly fitted. however, the output power for 1663 is 40w and the output power for the amp in its original design is 50w.
Hammond 1650RA (https://www.hammfg.com/files/parts/pdf/1650RA.pdf?v=1697661945) with 5k:4-8-16ohm,100 watt seems to be another option
my questions are:
1. any comparison between 1650RA or LL1663 for the amp?
2. any other recommendation for the OPT?
Thank you all in advance.
I want to do some mod to improve the sound quality of my mono-block amp, as attached.
As suggested, a change of high quality of OPT would be an option.
As shown, it is a PP with parallel EL34 for each path.
Lundahl LL1663 with 5k:8ohm would be possibly fitted. however, the output power for 1663 is 40w and the output power for the amp in its original design is 50w.
Hammond 1650RA (https://www.hammfg.com/files/parts/pdf/1650RA.pdf?v=1697661945) with 5k:4-8-16ohm,100 watt seems to be another option
my questions are:
1. any comparison between 1650RA or LL1663 for the amp?
2. any other recommendation for the OPT?
Thank you all in advance.
Last edited:
How about selling them and purchase something else ( A Dynaco MkIII for instance ? 60w less complex, perfectly balanced and good spare availability)
Have you measuired the original OPT and found it of low quality? Nothing says it should be of low quality.
Last edited:
@artosalo @revintage It works fine now. I just would like to have an improvement of the sound. Some said OPT, tubes, coupling cap are feasible upgrades.
It appears that the amp uses global feedback. If that's the case, then changing the output transformer will require some work with an oscilloscope to retune the feedback and make sure the amp is stable. I would not just drop a different OPT in there and assume the amp will be stable.
@grovergardner Thank you for your professional response. I am currently unskillful enough to handle that kind of mod, as you proposed. I may try to upgrade coupling cap first.
Post of picture of what it looks like under the hood. From what I remember these used pretty decent coupling capacitors in the first place. They were paper in oil type from Jensen if I remember correctly...
I will let you in on a little secret. If you run these PP amps with unmatched EL34's for some time, then the core's of the OPT's could become magnetized and the sound quality will suffer. These amps don't have any way to individually bias the output valves... They must be run with matched tubes if you want to avoid magnetizing the cores of the OPT's. Same goes for Jadis and others PP amps of this 'vintage'.
Other than that, the OPT's are really fantastic in these. They are the single best part of these amps. I wouldn't replace them with Lundahl.. certainly not Hammond...
I de-gaussed the OPT's and the sound quality was immediately back to 'good as new'. The difference was night and day better.
I will let you in on a little secret. If you run these PP amps with unmatched EL34's for some time, then the core's of the OPT's could become magnetized and the sound quality will suffer. These amps don't have any way to individually bias the output valves... They must be run with matched tubes if you want to avoid magnetizing the cores of the OPT's. Same goes for Jadis and others PP amps of this 'vintage'.
Other than that, the OPT's are really fantastic in these. They are the single best part of these amps. I wouldn't replace them with Lundahl.. certainly not Hammond...
I de-gaussed the OPT's and the sound quality was immediately back to 'good as new'. The difference was night and day better.
Last edited:
Triodes have relatively low inherent plate resistance. There is little benefit to load the plate with a current source (comparing to a simple plate resistor). You get diminishing of return. That’s why it is over complicated.
Last edited:
@jxdking - in SRPP the upper triode is not 'just' a current source.
I suggest this explanation of how SRPP works: https://ttradio.net/srpp-amplifier-theory-pictures/
They were quite popular when this amplifier was built. This amplifier sounds really great despite it's flaws (see my comments above on how to rectify them). One of the things that make this amplifier sound so good is it's very fine quality output transformers.
I suggest this explanation of how SRPP works: https://ttradio.net/srpp-amplifier-theory-pictures/
They were quite popular when this amplifier was built. This amplifier sounds really great despite it's flaws (see my comments above on how to rectify them). One of the things that make this amplifier sound so good is it's very fine quality output transformers.
Last edited:
Yea, that’s my thought experiment. If the upper tube has infinite gm, then it is a current source. In reality, the AC resistance is only couple times of the plate resistance, which is worse.
I spent couple minutes to review your link. I don’t agree the author calculated with Mu instead of gm. The amplification factor would be much worse if you have any load or plate resistor.
I see that you still don't understand this circuit. It's not like a MU-Follower or Constant Current Source. It's essentially a push-pull circuit.
Another very good explanation of the SRPP circuit comes from DiyAudio's very own Merlin Blencowe: https://www.valvewizard.co.uk/SRPP_Blencowe.pdf
John Broskie also explains it as well on his tubecad website in a few places, but in my humble opinion Merlin's explanation is better
You can also look up the original patents for this circuit from 1957 but it don't explain much.
There is a lot more to consider regarding valve/tube amplifiers than just 'amplification factor'. The advantage of the SRPP (if correctly implemented) is:
1. Relatively high gain
2. Low output impedance
3. Relatively low distortion
The problem with just using a resistive load is that the output impedance sometimes is higher than desired. Back in the 1990's they were doing SRPP all over the place. In the early 2000's the Chinese were copying it a lot too. Sometimes it was implemented in places where it was not needed (I agree with that).
This amp can be fixed without great expense. Changing the input/drive SRPP to something else won't improve it.
Another very good explanation of the SRPP circuit comes from DiyAudio's very own Merlin Blencowe: https://www.valvewizard.co.uk/SRPP_Blencowe.pdf
John Broskie also explains it as well on his tubecad website in a few places, but in my humble opinion Merlin's explanation is better
You can also look up the original patents for this circuit from 1957 but it don't explain much.
There is a lot more to consider regarding valve/tube amplifiers than just 'amplification factor'. The advantage of the SRPP (if correctly implemented) is:
1. Relatively high gain
2. Low output impedance
3. Relatively low distortion
The problem with just using a resistive load is that the output impedance sometimes is higher than desired. Back in the 1990's they were doing SRPP all over the place. In the early 2000's the Chinese were copying it a lot too. Sometimes it was implemented in places where it was not needed (I agree with that).
This amp can be fixed without great expense. Changing the input/drive SRPP to something else won't improve it.
Thanks. Very interesting reading.
There are still something that I don’t fully agree the conclusion.
Pro:
1. It is a push pull, (I can see that now). The current capability could be more than the double of the plate bias current. It might be able to work at class AB in some degree.
2. It has lower distortion comparing to single ended one.
3. The output impedance is halved, because of the push pull configuration.
Con:
1. The gain (without load) would be lower. It is especially true for the optimized version that the upper tube’s Ra equals to Lower tube’s Ra. The upper portion is far from a current source. Thus, the max gain would be lower.
2. The output impedance is still much higher than a cathode follower. I don’t think the paper got it right here. It might drive 15 Ohm headphones. It is because of the current capability of the push pull topology, not because of the low output impedance. For a 15 Ohm headphone, you don’t need more than 100ma at peak. You could get away with SRPP with 50mA bias (or less, driving into class AB).
My summary, SRPP is useful to double the current capability. It has relative low distortion with the same load. I still don’t agree to use it in the input stage or phase inverter stage. Those stages are light loaded (hundreds K). As said, you got diminishing of return.
Last edited:
I'm not a big fan on SRPP, but it was popular back when this amp was marketed.
As usual, the weakness in this amp's design is how it handles un-matched pentodes. This will over time, running the amp with unmatched pentodes leads to the OPT core becoming magnetized, which really hurts their performance and frequency response. But of course when these amps were new, nobody gave a thought to that right?
As usual, the weakness in this amp's design is how it handles un-matched pentodes. This will over time, running the amp with unmatched pentodes leads to the OPT core becoming magnetized, which really hurts their performance and frequency response. But of course when these amps were new, nobody gave a thought to that right?
@soulmerchant Could you please let me know how to de-magnetize the OPT? simply replacing with newly matched tubes?