Problem with SSTC

[Nunu00]

In the end I decided to redo the whole pcb in the smd version because the cost wasn't too much. As driver I will use the NCP81071 with 18V and 1:1:1 GDT.

In the meantime, however, I have redone some tests on the old board to try to understand something more now that I have a decent oscilloscope. First I tested all the passive components with the DMM and noticed that a diode in parallel with the 6.8 ohm resistor was shorted so I removed it.
Then I tested the various ICs individually and the 555 seems to work correctly:




On the inverter I connected a 100k resistor from pin 2 to the right pin of the 1k resistor and in these conditions on pin 1 I see this:



I'll have to try other resistors to find the correct frequency.


Then I applied to the antenna a square wave at about 200 kHz generated with a nodemcu and on pin 2 of the inverter I have:



and on pin 4 I have:



So this IC also seems to me to work fine.


Finally I tested the output of the MCP on the primary of the GDT and saw this:



and on a secondary I was seeing this:



I'm not very sure about these last signals but they seem ok to me anyway.

[davekni]

I'm not very sure about these last signals but they seem ok to me anyway.

Yes, I think signals are all good.  BTW, scope channel 2 is set for AC coupling.  I'd suggest setting scope channels to DC-coupled for almost all use.  There are some situations where AC-coupled is useful, such as measuring ripple of a DC supply.

I'll have to try other resistors to find the correct frequency.

Best to measure frequency with antenna connected and scope on 74HC14 pin 2 or 4.  Antenna adds capacitance as does scope probe, so real conditions are appropriate for setting frequency.  Doesn't hurt for frequency to be on the low side (high resistance).  Will still have edges to help start oscillation and keep 74HC14 pin 1 between its rising and falling threshold voltages.  A little high is OK.  To high can cause excess driver chip power dissipation in the event proper oscillation does not start.

With antenna connected, self-oscillation frequency will likely be noisy, as antenna picks up stray fields from line power wiring or whatever.  That is fine and normal.

[Nunu00]

Now I have a doubt regarding the choice of the resistor to add to the inverter because having to make the card in the smd version I believe that all the tests done on this old version are useless on the new one because change a lot of things.

Since I have to order the components, I think I should take a few resistors of different values but I have no idea which range to choose. Any suggestion?

[davekni]

Since I have to order the components, I think I should take a few resistors of different values but I have no idea which range to choose. Any suggestion?

Obviously these circuits work for many people with no resistor there, so don't get too concerned about value.  (Though startup issues do sometimes occur.)  A value on the high side such as 470k will be better than none, even if frequency ends up quite low.  Still keeps 74HC14 input close to threshold voltage.

Another option:  Most of my ECBs are surface-mount.  However, for values such as this where I may want to change values, I add two holes for a through-hole part, often on top of the SMD part.  Then I add 1-pin sockets to the two holes and plug in leaded parts.  You could have a high value such as 470k or even 1meg mounted to the SMD pads, then parallel with a leaded resistor if you want to match frequency better (ie. if you still have startup issues).

Yet another option:  If you construct many electronic projects, a resistor kit is quite handy, such as this one:
https://www.ebay.com/itm/263383970368
Select 1 set of "1206 SMD Resistor 1% (170..." options.  About $29 with shipping.  (Or smaller size parts if you want.)  This one is 25 parts per value.  For not much more cost you can find ones with 50 parts per value.  That is what I'm using.

[Nunu00]

After purchasing all the components in the smd version, redoing the schematic and the pcb with the suggested changes, the circuit doesn't work and this time I can't figure out what the problem might be. As you can see from the photos, the schematic has remained practically the same as the original one, I just added the additional resistance on the inverter and some decoupling capacitors. For the pcb I tried to follow the classic design rules and all the tracks should have been correctly sized.





As soon as I received all the material I started soldering the power supply part first, once that was done, I tested it and everything worked fine, all 3 voltages were present and there was no anomalous absorption.
Then I soldered the inverter part and that of the 555. Here the problems started because as soon as I switched on everything seemed to be fine and I started checking the output of the 555 with the oscilloscope. Something was wrong here as well as the output was very "ballie" and turning both control pots didn't change anything.

While I was checking the 555 the 7805 started to get very hot and after doing some measurements I found that the problem was the inverter that had started drawing too much current, so I left the 555 aside and concentrated on the inverter. I double-checked all the connections and the pcb tracks and I didn't find anything anomalous so I replaced the inverter with a new one, for the first 30/40 seconds after powering up the absorption was correct but shortly after the second inverter too it started having the same behavior as before.

Unfortunately I no longer have spare parts for that inverter and therefore I can no longer test that part. On the other hand, I have several spare parts for the 555 and therefore to remove any doubts I only soldered the 555 part on a new pcb board and the absurd thing is that in this case the 555 does not work at all, I find a dc voltage on the output.

I really lost hope, I don't know what to check anymore because everything seems correct to me but nothing works anyway.

[ZakW]

Hello Nunu00,

Sorry to hear that you have lost hope with your project. I have been there many times!

I did notice something in your schematic, it might not be the issues you reported at the moment but it might be worth looking into. Did you flip one of the outputs of the GDT? To me it looks like each gate is getting the same phase signal which will make both of them switch on at the same time and short out.

Good luck!

[Nunu00]

The gdt wasn't even connected, just to avoid burning some components I started testing the components individually.

[ZakW]

Then I soldered the inverter part and that of the 555. Here the problems started because as soon as I switched on everything seemed to be fine and I started checking the output of the 555 with the oscilloscope. Something was wrong here as well as the output was very "ballie" and turning both control pots didn't change anything.

By inverter do you mean gate drive chip (IC3, NCP8...)? Update: Nevermind, I read some of your most recent posts. I see that it is the inverter.
What do you mean by "output was very "ballie"?

While I was checking the 555 the 7805 started to get very hot and after doing some measurements I found that the problem was the inverter that had started drawing too much current, so I left the 555 aside and concentrated on the inverter. I double-checked all the connections and the pcb tracks and I didn't find anything anomalous so I replaced the inverter with a new one, for the first 30/40 seconds after powering up the absorption was correct but shortly after the second inverter too it started having the same behavior as before.

If I understand correctly that the inverter here is the IC3 chip they should never be powered on without anything connected to their output. I am not exactly sure what causes it but they have a tendency to burn up without being connected to anything. You mentioned that you did not install your GDT yet, that could be why they failed. *be sure to check GDT phasing in your schematic.

Unfortunately I no longer have spare parts for that inverter and therefore I can no longer test that part. On the other hand, I have several spare parts for the 555 and therefore to remove any doubts I only soldered the 555 part on a new pcb board and the absurd thing is that in this case the 555 does not work at all, I find a dc voltage on the output.

There may be an issue with the schematic/PCB. Are you able to breadboard the 555 timer interrupter section exactly like your schematic to see if it works?

[davekni]

If I understand correctly that the inverter here is the IC3 chip they should never be powered on without anything connected to their output. I am not exactly sure what causes it but they have a tendency to burn up without being connected to anything.

Unloaded outputs are not a problem unless there is some other issue too.  Unloaded inputs can be a problem, including antenna input.  It is possible for antenna input to couple to some other signal on board and oscillate at high frequency, causing excess power dissipation.  Best to test with antenna input grounded or connected to a signal generator.

555 issue will be easier to debug.  Use a scope to measure voltage on each pin.  If all are DC, record value for each pin.  Then looking at 555 data sheet, it is possible to determine if this is a valid state for 555.  If not (and no pin is shorted), then 555 chip must be bad.  If it is valid, then there is some issue with external circuit.

[Nunu00]

What do you mean by "output was very "ballie"?

I don't know how to explain it, but the output signal was not a constant square wave, it varied continuously both in frequency and in duty cycle.

If I understand correctly that the inverter here is the IC3 chip they should never be powered on without anything connected to their output. I am not exactly sure what causes it but they have a tendency to burn up without being connected to anything.

The IC3 chip was not present on the board, with the word inverter I mean the SN74 chip (IC5).
However I'll keep your suggestion in mind when I've solved this problem and move on to soldering the IC3 chip.

Are you able to breadboard the 555 timer interrupter section exactly like your schematic to see if it works?

All components are smd but I could try with some soldering work to test the 555 on breadboard.

Unloaded outputs are not a problem unless there is some other issue too.  Unloaded inputs can be a problem, including antenna input.  It is possible for antenna input to couple to some other signal on board and oscillate at high frequency, causing excess power dissipation.  Best to test with antenna input grounded or connected to a signal generator.

Unfortunately this was a detail that I didn't know, now I don't even have spare chips to try and before buying more I would like to try to solve the problem of the 555 chip.

555 issue will be easier to debug.  Use a scope to measure voltage on each pin.  If all are DC, record value for each pin.  Then looking at 555 data sheet, it is possible to determine if this is a valid state for 555.  If not (and no pin is shorted), then 555 chip must be bad.  If it is valid, then there is some issue with external circuit.

I did that and the voltage was DC on all pins, but from the datasheet I couldn't tell if it was a valid state or not. However on the pins I measured this:

1 -- 0 V
2 -- 0.6 V
3 -- 10.86 V
4 -- 12.11 V
5 --8.08 V
6 -- 0.6 V
7 -- 0 V
8 -- 12.11 V

[davekni]

I did that and the voltage was DC on all pins, but from the datasheet I couldn't tell if it was a valid state or not. However on the pins I measured this:
1 -- 0 V
2 -- 0.6 V
3 -- 10.86 V
4 -- 12.11 V
5 --8.08 V
6 -- 0.6 V
7 -- 0 V
8 -- 12.11 V

Yes, that is a valid 555 state.  Input (pins 2 and 6) is low so, causing output (pin 3) to be high.
So problem is most likely in parts around 555.  That high output should pass through POT2, D5, and R6 to cause a high on the input.  Measure voltage through those nodes to see where the high output becomes low.

[Nunu00]

Measure voltage through those nodes to see where the high output becomes low.

I just made the measurements, with POT2 set to minimum resistance on pin 3 node of the 555 chip I measure 10.77V, after POT2 I measure 10.76V, after D5 I measure 10.12V and after R6 I measure 0.7V.
From what I understand this voltage shouldn't be that low, but if I measure the resistance of R6 with the DMM I read about 2.1k which is correct.

[davekni]

after D5 I measure 10.12V and after R6 I measure 0.7V.

9.4V / 2.1k = 4.3mA.  So something is conducting an extra 4.3mA from 555 pin 2/6 to ground.  Could be a faulty C9, a bad 555 chip, or a short on your board to some other node.  If 555 is socketed, you could remove it, tie pin 3 to 12V, and measure voltages again.  If pins 2 and 6 are high, 555 is bad (pins 2 or 6 are conducting to ground).  I presume you have checked that 555 isn't inserted backwards (180 degrees rotated).

[Egg]

Kind of an old topic but i had the exact same problem with the 555 oscillator two times and I may have finally found the solution (at least in my case). The first time i got this i somehow fixed it by replacing components but did not know what did it. Now i had the problem again while soldering a new driver and after multiple hours of attempts i realized that the tantalum capacitor i was using was connected in wrong polarity. I did not even know that the cap was polarized since the legs are same lenght and there was only a very tiny cross marking.
So my guess is that the cap leaks some voltage to ground when connected in reverse polarity, but doesnt explode because it is rated for 35v and only 6-8v is coming from the pot. Im not sure if this is the right explanation but thats my guess.