Yup, however they are larger than the TPA v1.2 boards you have. I'd second Uriah's suggestion to try to repair the board - if the pad that delaminated has no trace connected to it on that side, there is no problem and electrical connectivity is the same as before. You can carefully solder the lead to the through-hole.
Alternatively, ask Uriah for a bare TPA v1.2 board if available.
You can check resistances in-circuit without desoldering - just compare with the working board to see if it looks similar.
Finally, check R11 to see if it is close to 1 ohm. This often fails open due to various human errors.
I only have two left and would like to keep them to build in the future. I have no modules built up or operating so these are truly my last boards from the buy.
With the bad one running you might place your hand flat a half inch or so above the tallest component so you can feel heat rising from the board. It will work. You can even feel the heat from an LED doing this. What I am thinking is you may have a short somewhere and this will usually make at least one part get way to hot. Then you can go to touching cases, but not metal parts to see if something is to darn hot. If your relay doesnt want to click then it has voltage cuz its closing a tiny bit but not all the way.
Look for solder flags on the bottom of the board. Soldermask is not as good as it should be with some board manufacturers. Sometimes you can put a microscopic scratch on the mask just by shipping two boards together. Then you solder and a little flag comes off of the through hole to that scratch and you have a short to the ground plane or another close trace. You will lose voltage here as there will now be a loss of current and it basically makes a voltage divider. I know you have probably reflowed the solder joints but its something I run into from time to time with boards built in China because the mask is not very good. My easy fix is to flux each joint and then heat it. Works every time. Rosin flux solder is not as good as real flux plus rosin core solder. If you dont clean your tip and remove excess solder then rosin core solder will get thick and sludgy and these flags are easy to make. Real flux often applied will remove this problem entirely. May not even be your problem, but have to check every possibility.
With the bad one running you might place your hand flat a half inch or so above the tallest component so you can feel heat rising from the board. It will work. You can even feel the heat from an LED doing this. What I am thinking is you may have a short somewhere and this will usually make at least one part get way to hot. Then you can go to touching cases, but not metal parts to see if something is to darn hot. If your relay doesnt want to click then it has voltage cuz its closing a tiny bit but not all the way.
Look for solder flags on the bottom of the board. Soldermask is not as good as it should be with some board manufacturers. Sometimes you can put a microscopic scratch on the mask just by shipping two boards together. Then you solder and a little flag comes off of the through hole to that scratch and you have a short to the ground plane or another close trace. You will lose voltage here as there will now be a loss of current and it basically makes a voltage divider. I know you have probably reflowed the solder joints but its something I run into from time to time with boards built in China because the mask is not very good. My easy fix is to flux each joint and then heat it. Works every time. Rosin flux solder is not as good as real flux plus rosin core solder. If you dont clean your tip and remove excess solder then rosin core solder will get thick and sludgy and these flags are easy to make. Real flux often applied will remove this problem entirely. May not even be your problem, but have to check every possibility.
Thanks all, and I'll PM you Bob about the board as I'll take it.
I want to try and keep to this version of the board as I've done the casing, plus I would rather start a fresh build to try and eliminate any user errors. I built the boards about 1.5 years ago, however only got around to finishing them now, and I've learnt a lot since then.
Udaily: I've already tried your heat checking method, although what I did was run a song through the working board, but with the "faulty" board connected to the power as well. I then disconnected the power and started by feeling the heat of the components; the LM3886 was hot (not connected to the heat sink for this short period), but that was the only thing that was hot.
From your discussion about the minute scratching, that could well be a possibility, and I'll take that into consideration next time.
I want to try and keep to this version of the board as I've done the casing, plus I would rather start a fresh build to try and eliminate any user errors. I built the boards about 1.5 years ago, however only got around to finishing them now, and I've learnt a lot since then.
Udaily: I've already tried your heat checking method, although what I did was run a song through the working board, but with the "faulty" board connected to the power as well. I then disconnected the power and started by feeling the heat of the components; the LM3886 was hot (not connected to the heat sink for this short period), but that was the only thing that was hot.
From your discussion about the minute scratching, that could well be a possibility, and I'll take that into consideration next time.
Last edited:
Thanks for the heads-up. Very interesting indeed. Problem is the LF01 still uses the LM318, which is the only component I can't source locally. Just looking for a direct drop-in alternative now.
It is your call, but linuxguru can send you a completed unit. That's what I bought. I'll just say it was a most significant upgrade to the MyRef series prior to Dario's Fremen Edition.
I'll also be happy to send you an LM138 via USPS if you wish. The shipping is generally between $4 - $6. PM me if I can help.
I'll also be happy to send you an LM138 via USPS if you wish. The shipping is generally between $4 - $6. PM me if I can help.
Bob - thanks for the plug. I now also have a fully-discrete LF05 prototype which also works instead of the LM318. I'll send out a pair with your LF03s, which you are welcome to evaluate at your convenience.
The TL071 will work in place of an LM318, with similar stability but worse sonics. The OPA627 and LME49710 are both marginally unstable in a Rev C - compensation changes will be required, and at the minimum C10 may have to be removed.
Yup, the LF05 is a direct drop-in replacement for the LM318, and the first pair of prototypes work without modification in a Rev E. I will get a few prototypes into local DIYers' hands to ensure that there is no regression before making it generally available - give it 4-8 weeks.
Preliminary audible sonics of the LF05 are comparable to the LF01.
Preliminary audible sonics of the LF05 are comparable to the LF01.
Both are correct.
The 33k value comes from LM3886 datasheet and Penasa's original BOM.
TP boards (including v1.2) also spefify the 10K value.
All my BOMs speficy 10K
Such value comes from a Nelson Pass post.
According Penasa the amp work as he intended with values around 1uF, TP boards specify no more than 2.2uF but I suppose it will work with higher values also.
But I suggest you to stay in the 1uF-1.5uF range
Thanks, Dario. I'll stick with the 1uf cap then.
Looks like the Omron relay is out of stock now. Would this one work just fine? 833H-1C-C-24VDC Song Chuan | Mouser
Looks like the Omron relay is out of stock now. Would this one work just fine? 833H-1C-C-24VDC Song Chuan | Mouser
It should but never tried it... or you can buy the non sealed Omron: 653-G5LE-1-DC24
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.