Joe's MPSA18 board kit
- Thread starter Skratch
- Start date
- Joined
- Jan 14, 2011
- Messages
- 74,248
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
Thanks for posting this Mark and creating awareness of this change. I am now matching all MPSA18s shipped with the PL14_20 kits within 200uV of each other on Vbe to achieve lower amp offset.
I will be experimenting with BC550B devices that I have ordered as well which pin out the same as the original TIS97 device.
Both MPSA18 and the BC550B are better devices than the TIS97 on paper and lower noise figure as well.
I will be experimenting with BC550B devices that I have ordered as well which pin out the same as the original TIS97 device.
Both MPSA18 and the BC550B are better devices than the TIS97 on paper and lower noise figure as well.
- Joined
- Jan 14, 2011
- Messages
- 74,248
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
- Joined
- Jan 14, 2011
- Messages
- 74,248
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
- Joined
- Jan 14, 2011
- Messages
- 74,248
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
Joe, have you considered slipping in two pots to set the DC Offset in the new board design similar to the 100K pots PL used in the PL 400C? Also, many of the bias adjustments I have made to the White Oak Audio Board seem to be "touchy." Would it be possible to re-source a quality 10 or 12-turn pot for the R20L and R20R pots? Maybe Mouser P/N 81-PV37P502C01B00 (this one is 6mm or 1/4") or something similar? The price just jumped on this one because the catalog lists $1.84 and now it is $2.35. I think everything is going up...
Hi Mark
You can put a 10 or 12 turn pot in the bias location but to a large extent that is overkill. The bias is touchy not because of the pot but because the 2N3403 is constantly moving at the time you are adjusting it (it is responding to the thermal changes on the backpanel as well as the internals of the 2N3403 b-e junction, which is its job). The bias is not all that critical as long as it is reasonably close. Lee can probably attest to that because he has been doing some distortion vs. bias setting work recently.
You can put a 10 or 12 turn pot in the bias location but to a large extent that is overkill. The bias is touchy not because of the pot but because the 2N3403 is constantly moving at the time you are adjusting it (it is responding to the thermal changes on the backpanel as well as the internals of the 2N3403 b-e junction, which is its job). The bias is not all that critical as long as it is reasonably close. Lee can probably attest to that because he has been doing some distortion vs. bias setting work recently.
- Joined
- Jun 12, 2010
- Messages
- 12,037
- Location
- OZONE ALLEY MARS (Visitor)
- Tagline
- Wasted Days and Wasted Nights
Hello Mark
The heatsink is not required on the TIP31C. It is loafing. The reason some have put the heatsink on the device is that those users are powering auxiliary devices like DC protect relays from that power source. It was designed to power the amp front end. When it gets used for other applications it heats up fast because it is a linear regulator with either 60 or 80 volts across it in the 400 and 700 applications respectively.
The TIP31C transistor is so large to handle the fast inrush current surge at power up time into the electrolytic caps in the power section. Those caps are large on purpose to provide hold up power on power down to minimize turn off thump
Hope this helps with the understanding
The heatsink is not required on the TIP31C. It is loafing. The reason some have put the heatsink on the device is that those users are powering auxiliary devices like DC protect relays from that power source. It was designed to power the amp front end. When it gets used for other applications it heats up fast because it is a linear regulator with either 60 or 80 volts across it in the 400 and 700 applications respectively.
The TIP31C transistor is so large to handle the fast inrush current surge at power up time into the electrolytic caps in the power section. Those caps are large on purpose to provide hold up power on power down to minimize turn off thump
Hope this helps with the understanding
Hi Mark L
I have resisted putting in a DC offset adjustment for several reasons (but I did think about it). Those are:
It adds a noise source to the signal path that is not compensated out by the amp circuitry.
It adds a reliability vulnerability to go along with the feature, I always weigh those trade offs. On 2 of the PL400s I have, the bias adjustment pots have gone bad
Transistor matching has other side benefits, more consistent side to side gain in the long tail pair
I have resisted putting in a DC offset adjustment for several reasons (but I did think about it). Those are:
It adds a noise source to the signal path that is not compensated out by the amp circuitry.
It adds a reliability vulnerability to go along with the feature, I always weigh those trade offs. On 2 of the PL400s I have, the bias adjustment pots have gone bad
Transistor matching has other side benefits, more consistent side to side gain in the long tail pair
- Joined
- Jan 14, 2011
- Messages
- 74,248
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
I have top agree on the matching as opposed to the pot. Matching may be labor intensive to begin with but the results are over time a lot more consistent without the additional attention. If you've ever had to do the offset on a cranky 700 Ser I with a PL 0171 board you would know what I'm talking about. The folks back in the day weren't kidding when they said "warm the amp" . My Spec 2 has to on for at least 45 minutes before you even think about it .
With Don's 400 I rteplaced the 3403's with the new 5088's. rock steady, good resolution and excellent range. Joe did good there.
Bias vs Distortion., We just put on a couple new wells we bought yesterday so I HAVE to go by band check on em, but I'll post some results on this.. Ed blackwood has donbe some EXTENSIVE work in this area. If he ever joins he could be a large contributor to this section.
With Don's 400 I rteplaced the 3403's with the new 5088's. rock steady, good resolution and excellent range. Joe did good there.
Bias vs Distortion., We just put on a couple new wells we bought yesterday so I HAVE to go by band check on em, but I'll post some results on this.. Ed blackwood has donbe some EXTENSIVE work in this area. If he ever joins he could be a large contributor to this section.
I was thinking about both transistor matching and the DC offset adjustment pot. My theory was that with the newer and better quality components (sealed pots vs open pots, for example) and the use of MPSA18 transistors for the long-tailed pair, that the noise and reliability issues would be less of a problem. I didn't think the PL implementation of the DC offset pot was in the signal path because it simply balances the bias of the input pair.
I noticed the BC550 and the MPSA18 were added to the last build sheet. Is there a preference? I see that the MPSA18 costs seven times as much as the BC550B, is it seven times as good?
I noticed the BC550 and the MPSA18 were added to the last build sheet. Is there a preference? I see that the MPSA18 costs seven times as much as the BC550B, is it seven times as good?
Hi Mark
The offset pot adds either plus or minus current into one of the inputs of the op amp that forms the front end. That current originates from the plus and minus power supplies within the amp. Unless perfectly clean (which it is not), it will be a noise source input into the amp and the amp will amplify it with the same gain as any other signal.
I am sampling the BC550B so I don't have any comment on these yet Mark until I receive the samples and verify them. On paper, they look a little better than the TIS97, all around. I would not infer anything from the pricing. The MPSA18 appears to be the best device sampled so far. There are more contenders out there as well. With the TIS97 unobtainium at this point, I had to look for viable alternatives.
The offset pot adds either plus or minus current into one of the inputs of the op amp that forms the front end. That current originates from the plus and minus power supplies within the amp. Unless perfectly clean (which it is not), it will be a noise source input into the amp and the amp will amplify it with the same gain as any other signal.
I am sampling the BC550B so I don't have any comment on these yet Mark until I receive the samples and verify them. On paper, they look a little better than the TIS97, all around. I would not infer anything from the pricing. The MPSA18 appears to be the best device sampled so far. There are more contenders out there as well. With the TIS97 unobtainium at this point, I had to look for viable alternatives.