I am inclined to suspect that Q203 has been compromised the top of C203 had a bulge in it and it was found to be leaky and it has been replaced. I think the most likely cause of the bulge would be down to a major failure such as the expiry of the output stage.
The polarity of the dc offset has not been specified however C203 has the negative plate connected to earth and is rated at 6V.
If the dc offset is negative that would not be healthy for the new C203 either - some designs include a parallel diode with reverse polarity as protection.
At the same time there is the CN206 bridge connection to eliminate as a possible contributory element.
The polarity of the dc offset has not been specified however C203 has the negative plate connected to earth and is rated at 6V.
If the dc offset is negative that would not be healthy for the new C203 either - some designs include a parallel diode with reverse polarity as protection.
At the same time there is the CN206 bridge connection to eliminate as a possible contributory element.
Hi guys! I really appreciate all your replies.
B-E voltages are identical on all four transistors (Q204, Q203, Q212 and Q222), at 590mV.
I guess we can exclude leakage through C202 also because the offset is towards the positive rail (about +145mV). Anyway, I had already gone through all the network between CN201 and Q204 with my multimeter and everything ohms out the same as on the other channel.
I have been looking for the connector myself, but this A300 doesn't seem to have it. Neither does it have a switch to put it in bridge mode. Anyway, I have established that the channels work independently - the balance control does what it's supposed to do.
I have checked these components. What I have found is that the base and emitter voltages on Q205 are identical to those on the other channel, however the collector voltage is higher with respect to the other channel, by an amount equal to the DC voltage I am seeing on the speaker terminals. Same thing happens on Q206: emitter and base voltages same as on the other channel, collector voltage higher by about 145mV.
The base voltages on the two darlingtons are also offset by about 145mV towards the positive rail, with respect to those on the other channel. The darlington are not at fault, since I have seen the same offset with the darlingtons out of circuit.
You are right about how C203 failed, because with the faulty darlingtons in circuit, the voltage at the emitters was about -2.2V. With the new darlingtons in, it is around +145mV, so the new cap should be OK.
I have checked Q203 (in circuit) with the transistor checker (Peak atlas DCA). It measured the same as the corresponding transistor on the other channel. I guess this excludes Q203 as being faulty... I have done the same with Q204, Q205 and Q206. They all measure the same as on the other channel - even the hfe is very close, to within a few units...
Additionally, I suspected leakage through C239, but pulling it out of the circuit made no difference.
Doesn't look like I am getting any closer to repairing this...
B-E voltages are identical on all four transistors (Q204, Q203, Q212 and Q222), at 590mV.
I guess we can exclude leakage through C202 also because the offset is towards the positive rail (about +145mV). Anyway, I had already gone through all the network between CN201 and Q204 with my multimeter and everything ohms out the same as on the other channel.
I have been looking for the connector myself, but this A300 doesn't seem to have it. Neither does it have a switch to put it in bridge mode. Anyway, I have established that the channels work independently - the balance control does what it's supposed to do.
I have checked these components. What I have found is that the base and emitter voltages on Q205 are identical to those on the other channel, however the collector voltage is higher with respect to the other channel, by an amount equal to the DC voltage I am seeing on the speaker terminals. Same thing happens on Q206: emitter and base voltages same as on the other channel, collector voltage higher by about 145mV.
The base voltages on the two darlingtons are also offset by about 145mV towards the positive rail, with respect to those on the other channel. The darlington are not at fault, since I have seen the same offset with the darlingtons out of circuit.
You are right about how C203 failed, because with the faulty darlingtons in circuit, the voltage at the emitters was about -2.2V. With the new darlingtons in, it is around +145mV, so the new cap should be OK.
I have checked Q203 (in circuit) with the transistor checker (Peak atlas DCA). It measured the same as the corresponding transistor on the other channel. I guess this excludes Q203 as being faulty... I have done the same with Q204, Q205 and Q206. They all measure the same as on the other channel - even the hfe is very close, to within a few units...
Additionally, I suspected leakage through C239, but pulling it out of the circuit made no difference.
Doesn't look like I am getting any closer to repairing this...
Last edited:
The transistor emitter currents can be deduced by Ohms law by dividing these voltage figures by the common value of 15R.
For the non-working channel the input transistor Q204 connects to R206 which drops 12.51mV /15 =0.834 m.a.
The nfb return transistor Q203 connects to R205 which drops 12.22mV/15 =0.8146 m.a.
The input transistor Q206 Ie outweighs Q205 Ie.
Looking to the working channel the input transistor Q212 connects to R244 which drops 11.88mV/15=0.792 m.a.
The nfb transistor Q222 connects to R249 which drops 12.69mV/15=0.846 m.a.
The input transistor Q212 Ie is outweighed by Q222 Ie.
My thoughts about Q203 having been compromised were not about complete failure but a degradation of the beta. You could use your Atlas tester to check this out.
To investigate further than that the voltage drops across R250,251, and 265 should be measured and compared to those in the working channel.
Last edited:
Hi mjona, thanks again.
Q203 has an hfe of 137, compared to 134 on Q222.
In fact, I have done an in-circuit test of all transistors with the Atlas. They all read very similar to the ones on the good channel, to within a few units of hfe and a few mV of B-E voltage. The only transistors I could not check were Q204 and Q212 (measured as series diode network) as well as Q209 and Q217 (measured as unknown/faulty component), but I guess that's because of the surrounding components.
The voltages on the resistors are as follows: 64.90mV on R221, 65.63mV on R220 and 4.40mV on R222, vs. 65.85mV on R250, 63.70mV on R251 and 4.23mV on R256.
Q203 has an hfe of 137, compared to 134 on Q222.
In fact, I have done an in-circuit test of all transistors with the Atlas. They all read very similar to the ones on the good channel, to within a few units of hfe and a few mV of B-E voltage. The only transistors I could not check were Q204 and Q212 (measured as series diode network) as well as Q209 and Q217 (measured as unknown/faulty component), but I guess that's because of the surrounding components.
The voltages on the resistors are as follows: 64.90mV on R221, 65.63mV on R220 and 4.40mV on R222, vs. 65.85mV on R250, 63.70mV on R251 and 4.23mV on R256.
The Ie of Q204 and Q203 as calculated from the voltage drops across 15R are different at 0.834 m.a. and 0.8146 m.a. If the hfe of these transistors is close the base currents should be similar in value flowing to 0 volts via 33k resistors R202 and R210. The voltage drop across these should also be close.
If you reverse the deployment of the transistor designate Q203 so it becomes Q204 and vice versa does that change the dc offset. If not an improvement we will have to look at other things.
If you reverse the deployment of the transistor designate Q203 so it becomes Q204 and vice versa does that change the dc offset. If not an improvement we will have to look at other things.
I have followed your suggestions to use base to emitter resistors. Thanks again for suggesting it.
The voltage across the output terminals is now 145mV. I have checked voltages on all transistors, and they are mostly the same, compared to the working channel, with only a few deviating by no more than 200mV.
Would this be a pass mark?
Did you revert the values to 47R after fitting the new Darlingtons.
It is possible one of the 100R test condition substitutes has changed in value due to the fault condition - such as if there was a short circuit across the internal base emitter junction in these packages after either of the internal emitter resistors in parallel or in series had failed.
Hi guys. I am glad to say that the Cambridge Audio A300 is fixed
The troublemaker was Q204. It had an unusually low hfe, of 115 (I tested a couple of new transistors and they all had an hfe of between 175 and 200). At the same time, Q203 had a rather high hfe, of 202 (different to the 137 value I got when I tested it in circuit). I suppose this imbalance was responsible for the DC offset at the output.
Now the offset is around -9mV. Hope it's not going to affect C203 too much. The other channel has a DC offset of about 10mV.
Thanks a lot mjona for identifying the issue was with either Q203 or Q204!
I have left the 47R resistors in circuit (and they tested fine). After removing the old darlingtons, I just soldered in 330R resistors from base to emitter, to test the rest of the circuit. I removed these resistors when I put the new darlingtons in
By the way, if you are wondering why 330R, I calculated the value using the working channel's base current for the darlingtons (voltage drop across the base resistors of 47R) and the B-E voltage. The B-E equivalent resistance of the darlingtons amounted to about 330 ohms.
Thanks a lot for all the support I got on the forum. I would have really struggled to fix this on my own...
The troublemaker was Q204. It had an unusually low hfe, of 115 (I tested a couple of new transistors and they all had an hfe of between 175 and 200). At the same time, Q203 had a rather high hfe, of 202 (different to the 137 value I got when I tested it in circuit). I suppose this imbalance was responsible for the DC offset at the output.
Now the offset is around -9mV. Hope it's not going to affect C203 too much. The other channel has a DC offset of about 10mV.
Thanks a lot mjona for identifying the issue was with either Q203 or Q204!
I have left the 47R resistors in circuit (and they tested fine). After removing the old darlingtons, I just soldered in 330R resistors from base to emitter, to test the rest of the circuit. I removed these resistors when I put the new darlingtons in
By the way, if you are wondering why 330R, I calculated the value using the working channel's base current for the darlingtons (voltage drop across the base resistors of 47R) and the B-E voltage. The B-E equivalent resistance of the darlingtons amounted to about 330 ohms.
Thanks a lot for all the support I got on the forum. I would have really struggled to fix this on my own...
Last edited:
In post 37 I suggested some leakage current to earth via C202, the only other route for current to explain the higher level of Ie is due to internal leakage in the transistor which increased Ic but decreased hfe.
There is diffusion capacitance between the base and emitter and between the base and collector.
The leakage could have occurred across those due to the trauma to C203. While the diode junction stayed intact.
The width of the collector could have reduced permanently in the process and the relationship between the base would have made the base relatively larger which would explain the reduction in hfe.
There is diffusion capacitance between the base and emitter and between the base and collector.
The leakage could have occurred across those due to the trauma to C203. While the diode junction stayed intact.
The width of the collector could have reduced permanently in the process and the relationship between the base would have made the base relatively larger which would explain the reduction in hfe.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.