Here is some measured volatges:
Does it look like the protection mode? And how could I by pass it?
Attachments
Maybe it is. Could you measure the voltage between R14 - R15? To deactivate protection just shorten R13 and R113 as Q7 if conducts disables serial mosfet.
Q8 and Q9(same on other side) are kind of SCR as ones triggered stay conducting. Trigger can come from Q6(current sensing) Q11//Q12 DC offset sensing and from D5 - D105 unequal output voltage sensing. The protection ones triggered can be reset with power switch off. Pay attention that the capacitors from the input power board should discharge and that could take time as PS regulator is off and no load on those capacitor. To make it faster put some load on those capacitors or on the PS regulator voltage inputs.
It is 28.15V when normal and about 0.633V on 7V mode.
- Turn off and wait for capactitor discharge seem bring back normal mode.
- Turn on and with DMM connected and do nothing, the 7V mode won't happen.
- And if I do some measure on board, 7V mode will appear.
R9 / R19 volage seem ok so it is not real overload? May be I will test with protection by pass on: both rail -> ok? -> one rail -> to see where it happen.
- Turn off and wait for capactitor discharge seem bring back normal mode.
- Turn on and with DMM connected and do nothing, the 7V mode won't happen.
- And if I do some measure on board, 7V mode will appear.
R9 / R19 volage seem ok so it is not real overload? May be I will test with protection by pass on: both rail -> ok? -> one rail -> to see where it happen.
Normally SP protection input if not connected should not trigger the protection, but just in case connect the SP inputs to the ground.
I did it. It seem the overload trigger instant when I turn on board with my dummy load (higher than before). Both channels were the same. Connect SP to ground didn't change anything. So I think I need to check things arround Q6 / Q106. I think it check overload too soon. In My PCB , I use 2N5401 istead of BC556B and 2N5551 instead of BC546B, is it ok?
It should be OK. I have 3k3 resistor in series with D1,D2,D3 and off cause the same with D101,D102, D103 in my BJT working regulator. You can add those too, but that does not explain so sensitive trigger, maybe you have a layout error or wrong resistors value somewhere. It should trigger on the load current higher than 10 A. Try to check if both sides react the same. It's enough that one side triggers and take other side down too. Make one side passive and check the trigger current level, and than other side the same.
I found a short at B-C of Q109 which cause by left over solder paste. I corrected that and remove the short at R13 and R113 ...
It run
-> Doesn't have weird overload too soon.
-> Status led display correct.
but have a strange behavier
-> If I turn PS Regulator on and then connect my DMM to GND / V+ / V- it make the PSU Regulator go to protection mode.
-> If I turn on PS Regulator but with my DMM already connected, it seem ok and I can move one point to test voltage in other place. But if I disconnect two point and then connect again -> protection mode.
It run
-> Doesn't have weird overload too soon.
-> Status led display correct.
but have a strange behavier
-> If I turn PS Regulator on and then connect my DMM to GND / V+ / V- it make the PSU Regulator go to protection mode.
-> If I turn on PS Regulator but with my DMM already connected, it seem ok and I can move one point to test voltage in other place. But if I disconnect two point and then connect again -> protection mode.
As I said before, try to find if both sides trigger protection mode, by disabling one side and test other and vice versa.
It triggered with 2 ohms loads. Both side. And if one side down, it also took other side down, too.
Were there any changes needed to the boards? I would like to order some. If they are good as is I will place an order.
Thanks, Terry
I think so but it has strange behavier which active protection mode when I touch any wire to V+ / V- or ground? I don't know if it is an error?
That sensitivity, I never had with my BJT PS Regulator, it's strange. Did you insert 3k3 resistor in series with the diodes in each overcurrent sensors?
No. I didn't. But when I added 3k3 to both side, the ultra sensitivity seem gone, I can use my DMM to measure things. I will test more to make sure everything is fine.
Could I ask how the current limit work?
Uhm, I think I'm wrong! It still happen sensitivity is less but still there, It not happen directly but after some measure, it still happen. Like measure negative rail then quickly measure positive rail.
Will increase 3K3 to 10K or greater will reduce this sensitivity?
Sorry for multiple posting. I think sensitive seem work fine after 3K3 added with one strange case:
- One stick of DMM to Vin- and other connect to Vout+
- One stick of DMM to Vin+ and other connect to Vout-
Two actions will definity make the board switch to protection mode (both rail). It never happen on real life usage so it seem unimportant.
But when I careful testing the overload and speaker protection (I only check the status led before - sorry):
- Both speaker protection and overload only bring the fault rail down but leave other rail unchange.
So could I ask where I can check the the interaction between two rails?
I used 20V zener for 22V and 12V zener for all 10V zener in schematic. Could it be a problem?
- One stick of DMM to Vin- and other connect to Vout+
- One stick of DMM to Vin+ and other connect to Vout-
Two actions will definity make the board switch to protection mode (both rail). It never happen on real life usage so it seem unimportant.
But when I careful testing the overload and speaker protection (I only check the status led before - sorry):
- Both speaker protection and overload only bring the fault rail down but leave other rail unchange.
So could I ask where I can check the the interaction between two rails?
I used 20V zener for 22V and 12V zener for all 10V zener in schematic. Could it be a problem?
Voltage drop on R9 caused by the current will start to open Q6. Then the current trough R8 will trigger "thyristor" ( C8, C9). There is a cap ( C3) parallel to R8 and it act as a time delay dependent on the current trough R8 (and caused by current trough R9), higher current - sorter time reaction (that time is in msec).
You can try with 10k but I don't think that is the reason for such high sensitivity. In my BJT regulator I did not notice anything similar even without that resistor. I added that resistor as a measure of the Q6 protection in case of very high current pulse.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.