Hi
Lets compare some output pins with the PRY output pin. You can read the resistances for pins 8 thru 12. Read the resistances and see if they compare with the PRY pin 115 resistance.
Lets compare some output pins with the PRY output pin. You can read the resistances for pins 8 thru 12. Read the resistances and see if they compare with the PRY pin 115 resistance.
Hi,
Let's do one more test. Apply power to the amp. Wait for 2 or 3 minutes. This is to allow the capacitors for +5M1 and +5M2 to charge up. Turn the power switch ON and then do a hard reset by grounding the reset pin and see what happen. It is possible that the signal Model that it is set high by +5M2 and if +5M2 take to long to charge up the capacitors at reset time it will select the wrong model number. By doing in this way are going to make sure that it will select the correct model number pin 17 = model detect(0=RX-V3000 1= DSP-AX2).
Let's do one more test. Apply power to the amp. Wait for 2 or 3 minutes. This is to allow the capacitors for +5M1 and +5M2 to charge up. Turn the power switch ON and then do a hard reset by grounding the reset pin and see what happen. It is possible that the signal Model that it is set high by +5M2 and if +5M2 take to long to charge up the capacitors at reset time it will select the wrong model number. By doing in this way are going to make sure that it will select the correct model number pin 17 = model detect(0=RX-V3000 1= DSP-AX2).
First tinfoil and heating attempt was not successfull (supercap removed). Tinfoil - uP pins contact was not really good so I want to try this one more time today. I also have a hot air solder with digital temperature adjustment and can go a bit higher than 80c.
After ~20 hours the supercap had 0.7V only...Mooly said:
tauro0221 said:
Thanks, will check that later as well. Now the board is completely out of chasis.tauro0221 said:
That sound like either the cap is faulty or something is drawing current from it (the uP ?)
Try taking the cap out completely, discharge all the rails down to zero, and then try it. You can fit a 100uF electrolytic in place of the super cap as a test.
Hi,
There are 2 ways that the capacitors voltage can be discharge. 1: The use of diode D514 and D515 to isolate the regulators from the capacitors. They prevented the regulators from discharging the capacitors . Number 2 is that normally to reduce power usage the micro it is put to sleep mode. When the micro it is in the sleep mode is when the power consumption will be at the minimum state. While the micro is in the sleep mode the only way to wake it up is by an interrupt. That's why the power switch it is connected to an interrupt. When you turn ON the power switch the micro will wake up. At the stage we do not what state is the micro and that's why we see the capacitors discharge rate it is too high. Also normally the super cap used are at least 1uf 5 volts and used to keep the setup data in the memory.
There are 2 ways that the capacitors voltage can be discharge. 1: The use of diode D514 and D515 to isolate the regulators from the capacitors. They prevented the regulators from discharging the capacitors . Number 2 is that normally to reduce power usage the micro it is put to sleep mode. When the micro it is in the sleep mode is when the power consumption will be at the minimum state. While the micro is in the sleep mode the only way to wake it up is by an interrupt. That's why the power switch it is connected to an interrupt. When you turn ON the power switch the micro will wake up. At the stage we do not what state is the micro and that's why we see the capacitors discharge rate it is too high. Also normally the super cap used are at least 1uf 5 volts and used to keep the setup data in the memory.
Supercap is removed. Tried to remove charge from +5M1 and power on - nothing. Dicharge everything + tinfoil trick - still nothing 🙁
Pin 8-12 does not short to a ground (no such a small resistatnce like PRY - MG).
Pin 17 MODEL is always high (around +5V). Hard reset while power button press also does not get the unit up.
Those two are always about +5V:
Pin 52 /RDY
Pin 54 /HOLD
From manual:
Pin 8-12 does not short to a ground (no such a small resistatnce like PRY - MG).
Pin 17 MODEL is always high (around +5V). Hard reset while power button press also does not get the unit up.
Those two are always about +5V:
Pin 52 /RDY
Pin 54 /HOLD
From manual:
Last edited:
Replaced resonator an supercap more out of desperacy, no changes.
What do you guys think on feeding +MB from somewhere else e.g. PC PSU? Mine 16-18V looks suspicious to me...
What do you guys think on feeding +MB from somewhere else e.g. PC PSU? Mine 16-18V looks suspicious to me...
Is the replacement supercap exhibiting a similar discharge rate to that of the previous one?
Well, if the 5V rails are unaffected, using a different supply will make little no difference. The regulators are rated at 35V input maximum.
I don't think adding a separate supply is a good idea either. If the voltage is good and the ripple acceptable then its fine.
@ r0k1e:
If you wanted to confirm the level of ripple on the 5V rails feeding the CPU, you can use your oscilloscope set to AC mode with a sensitivity of around 2mV per division. Ensure that the 'scope probe ground is located close to the measurement point in order to minimise inductive pickup. This will allow you to obtain a peak-to-peak measurement. I would suggest you pick up the 5V rail close to the regulator.
If you wanted to confirm the level of ripple on the 5V rails feeding the CPU, you can use your oscilloscope set to AC mode with a sensitivity of around 2mV per division. Ensure that the 'scope probe ground is located close to the measurement point in order to minimise inductive pickup. This will allow you to obtain a peak-to-peak measurement. I would suggest you pick up the 5V rail close to the regulator.
Thanks for clarification on +MB.
0s 5.01
20s 2.80
1m 20s 1.62
2m 20s 1.32
3m 20s 1.15
As for 5V ripple, I measured it right on 5V regulator and saw no ripple at 0.01V per division (that's the best resolution from my scope).
With replaced supercap I see even faster discharge, but this should be ok as I installed 0.22F instead old 1F (original is only 0.047F):currentflow said:
0s 5.01
20s 2.80
1m 20s 1.62
2m 20s 1.32
3m 20s 1.15
As for 5V ripple, I measured it right on 5V regulator and saw no ripple at 0.01V per division (that's the best resolution from my scope).
You need to find why that cap is discharging so quickly. It should retain its voltage fro weeks and months. When measuring the voltage do not leave a meter connected as the internal resistance of the meter will pull the voltage down. Just measure, get a reading and then remove.
There is a resistor R702 (2.2 ohm) going to pin 16 of the uP. It would be worth monitoring the volt drop across this (mv range) and seeing if the voltage falls to zero on power off.
There is a resistor R702 (2.2 ohm) going to pin 16 of the uP. It would be worth monitoring the volt drop across this (mv range) and seeing if the voltage falls to zero on power off.
An explanation as to why the CPU is consuming too much power is likely to be because it is not entering sleep mode when the amplifier is powered-down. In fact, it is becoming more apparant that it is not running the correct firmware contained in the ROM as the following behavourial observations tend to suggest:
1) The PDT line, which should be configured as an input, is being switched to that of an output and is being driven to logic 'low'. (Post #11)
2) The 16MHz clock (from the video) appears to halt, under software control.
3) The bus connections between the ROM and CPU appear to exhibit normal activity.
As I stated previously, the CPU either cannot properly access the firmware held in the ROM, or the ROM contents have changed, rendering the combination inoperative.
From the datasheet, the ROM (Fujitsu Semiconductor MBM29F400BC-70) has the ability to be programmed in-circuit, via a bootloader:
This device is designed to be programmed in-system with the standard system 5.0 V VCC supply. 12.0 V VPP is not required for write or erase operations. The devices can also be reprogrammed in standard EPROM programmers.
See datasheet
Perhaps it is now worth exploring the possibility of whether the ROM can be re-programmed, possibly by a third party. Alternatively, source a replacement PCB containing a compatible firmware version.
Another possibility, is that the CPU is still in its 'stuck' state. However, because it is running some code and switching clock and PDT lines, it seems less likely.
Assuming that the ROM and CPU are working perfectly (just not as a team), another reason for this behaviour could be explained by shorts between adjacent address lines or data bus lines.
1) The PDT line, which should be configured as an input, is being switched to that of an output and is being driven to logic 'low'. (Post #11)
2) The 16MHz clock (from the video) appears to halt, under software control.
3) The bus connections between the ROM and CPU appear to exhibit normal activity.
As I stated previously, the CPU either cannot properly access the firmware held in the ROM, or the ROM contents have changed, rendering the combination inoperative.
From the datasheet, the ROM (Fujitsu Semiconductor MBM29F400BC-70) has the ability to be programmed in-circuit, via a bootloader:
This device is designed to be programmed in-system with the standard system 5.0 V VCC supply. 12.0 V VPP is not required for write or erase operations. The devices can also be reprogrammed in standard EPROM programmers.
See datasheet
Perhaps it is now worth exploring the possibility of whether the ROM can be re-programmed, possibly by a third party. Alternatively, source a replacement PCB containing a compatible firmware version.
Another possibility, is that the CPU is still in its 'stuck' state. However, because it is running some code and switching clock and PDT lines, it seems less likely.
Assuming that the ROM and CPU are working perfectly (just not as a team), another reason for this behaviour could be explained by shorts between adjacent address lines or data bus lines.
Thanks guys, without your help I'd be stuck at the very beginning!
Finally some progress!
I found out that supercap fault with electrolyte leakage causing shorts is something common and start exploring of supercap area for hidden faults. I noticed a small piece of PCB was damaged and then restored by previous repair (several small rail parts). Also I realized that uP solder joints were refreshed or even whole chip replaced. Finally, I've found absolutely non-visible break in D9 data line (currentflow mentioned such possibility multiple times actually 🙂). Fixed that, assembled everything back - still not powering on 🙁 But! Now supercap charge is not draining that fast. To my regret I lost concentration and accidentally shorted 5V regulator output and COM pin
As a result, D509 1ss133 diode died. I replaced it with 1N4148 and then used some compressed air to blow out flux-alike substance from under ROM and uP chips to ensure this doesn't cause any shorts. So now having following symptoms: after assembling the board back and several first power-on attempts I saw similar to previous behavior: oscillator stops after some time. However, on probably 3rd-4th power on, oscillator worked continuously and PDT remained high all the time! Now I can't repro oscillator stop and have supercap voltage drain slowly. Also +MB is always about 17-18V. Q506 pins all are around +17V too. Other voltages like +5M1, +5M2, /RES, seem similar to what I had before. Something is still missing. Now I have a hope that ROM data is actually ok (or at least was ok until I started my repair 🙂).
Probably unrelated, but R704 in-circuit resistance is 2.2K instead of 2.7K. Should I pay attention to that (consequence of voltage regulator short may be)?
Finally some progress!
I found out that supercap fault with electrolyte leakage causing shorts is something common and start exploring of supercap area for hidden faults. I noticed a small piece of PCB was damaged and then restored by previous repair (several small rail parts). Also I realized that uP solder joints were refreshed or even whole chip replaced. Finally, I've found absolutely non-visible break in D9 data line (currentflow mentioned such possibility multiple times actually 🙂). Fixed that, assembled everything back - still not powering on 🙁 But! Now supercap charge is not draining that fast. To my regret I lost concentration and accidentally shorted 5V regulator output and COM pin
As a result, D509 1ss133 diode died. I replaced it with 1N4148 and then used some compressed air to blow out flux-alike substance from under ROM and uP chips to ensure this doesn't cause any shorts. So now having following symptoms: after assembling the board back and several first power-on attempts I saw similar to previous behavior: oscillator stops after some time. However, on probably 3rd-4th power on, oscillator worked continuously and PDT remained high all the time! Now I can't repro oscillator stop and have supercap voltage drain slowly. Also +MB is always about 17-18V. Q506 pins all are around +17V too. Other voltages like +5M1, +5M2, /RES, seem similar to what I had before. Something is still missing. Now I have a hope that ROM data is actually ok (or at least was ok until I started my repair 🙂).Probably unrelated, but R704 in-circuit resistance is 2.2K instead of 2.7K. Should I pay attention to that (consequence of voltage regulator short may be)?
I'm pleased to hear you've been making good progress!Thanks guys, without your help I'd be stuck at the very beginning!
Finally some progress!
I found out that supercap fault with electrolyte leakage causing shorts is something common and start exploring of supercap area for hidden faults. I noticed a small piece of PCB was damaged and then restored by previous repair (several small rail parts). Also I realized that uP solder joints were refreshed or even whole chip replaced. Finally, I've found absolutely non-visible break in D9 data line (currentflow mentioned such possibility multiple times actually 🙂). Fixed that, assembled everything back - still not powering on 🙁 But! Now supercap charge is not draining that fast. To my regret I lost concentration and accidentally shorted 5V regulator output and COM pin
Probably unrelated, but R704 in-circuit resistance is 2.2K instead of 2.7K. Should I pay attention to that (consequence of voltage regulator short may be)?
I would continue to check for broken address and data lines. My recommendation would be to make up some probes using sewing needles or pins with small, sharp points and to probe between the ROM and CPU directly on their pins, using a continuity tester. This also checks for invisible bad solder joints under the IC legs. Before you do this, switch off (obviously!) and discharge the supercap with a resistor for a minute or so as before. If it helps, you could place some masking tape adjacent to the IC package (or on its top surface) where you're testing and write a small dot against the pin when its test is completed.
The (high) voltage level on +MB is possibly caused by a faulty D315 zener diode on the sub power supply, as the circuit around Q312, Q313 and Q314 is responsible for regulating this rail. In post #1, you measured +MB at 13.79V and now it has increased?
R704 is unlikely to have suffered, as long as it hasn't vapourised. Its value is likely to read lower than 2K7 as it is being measured in-circuit.
Keep up the good work 🙂
Last edited:
Thank you! +MB indeed settled at 13.79V after oscillator stops. Last week I also measured it right from power on moment when oscillations still occurred: first it was 17.7V, then it changes in steps to 16.5, then back to 17, then about 14.5 and so on until it finally was settled down to 13.7 with PDT 0.13.
I haven’t yet checked if step-change behavior remained after D9 rail fix, but I’m sure +MB is closer to 17-18V range now.
I haven’t yet checked if step-change behavior remained after D9 rail fix, but I’m sure +MB is closer to 17-18V range now.
D315 on the sub PSU should have ~11V DC across it. Such voltage fluctuations would suggest that the +MB discrete regulator isn't regulating well under varying loads. But you can leave that until later (if you prefer), as it shouldn't affect the CPU's own 5V regulator (as long as it remains cool).
With D9 repaired all the uP <-> Flash ROM rails are good. Sewing needles method was really helpfull! I was too optimistic when I found that D9 line broken and expected the unit to be fixed soon but oh well...
Now the main board is removed again and I'm going to check the rest of uP pins and probably keep the board outside for a while to try complete discharge again.
No other damage I could find in supercap area. The uP is quite close, but see no problem that I can recognize.
P.S. did all the recent experiments with supercap removed.
Now the main board is removed again and I'm going to check the rest of uP pins and probably keep the board outside for a while to try complete discharge again.
No other damage I could find in supercap area. The uP is quite close, but see no problem that I can recognize.
P.S. did all the recent experiments with supercap removed.
- Status
- Not open for further replies.