Hello guys,
First off, I am building a 01A preamp base on Ale Moglia's (Bartola Valves)design. The schematic is this:
The only difference I made is to replace the 47K resistor to a 100k pot(volume control).
For power supply, I am using the same PSU I have been using, with no problems whatsoever, with my Type 26 preamp in a two chassis set up with HV from a tube rectified(AZ1)arrangement, plus two filament regulators(Rod Coleman's) using 16Vac transformer per to each set of diode rectifiers.
An umbilical cord is connected to the PSU to the main preamp chassis. I am getting 202V B+(LCLCLC) and 25V DC for each filament supply measured at the iNPUT of the Rod Coleman boards.
Here is the problem. Looking back at the schematic, I can measure 4,8V at the top of the filament resistor and 8.11 V at the cathode of the 201A, and 118 at the plate with respect to ground(0 V) and 30 mV across the resistor R5. If I flip the SW1, it takes the tube rectifier out with a flash. If I connect an interconnect at the output of the preamp to an input of say another amp who's Ground(0V) is connected to chassis ground(IEC ground), it takes the rectifier again. In other words, as soon as I have the chassis ground (IEC) connected to my 0V, it destroys the rectifier!
Also, the 4.85V is the minimum I can get from the Coleman Reg adjustment (using 3.9 ohms R1) and measuring from the Filament Input(+ and -) on the board, I am getting 25Vdc! Why is it too high?
Chassis is aluminum. Heat sinking is to the chassis, umbilical cord is a Neutrik 8 Pole with 1+ and1- as Filament Supply 1, 2+ and 2- as B+ and 0V, 3+ and 3- as HV 0V and IEC safety ground, and 4+ and 4- as Filament Supply 2.
Trying to seek help as I lost three AZ1's already and this project is getting too expensive
Thanks!
Abe
First off, I am building a 01A preamp base on Ale Moglia's (Bartola Valves)design. The schematic is this:
The only difference I made is to replace the 47K resistor to a 100k pot(volume control).
For power supply, I am using the same PSU I have been using, with no problems whatsoever, with my Type 26 preamp in a two chassis set up with HV from a tube rectified(AZ1)arrangement, plus two filament regulators(Rod Coleman's) using 16Vac transformer per to each set of diode rectifiers.
An umbilical cord is connected to the PSU to the main preamp chassis. I am getting 202V B+(LCLCLC) and 25V DC for each filament supply measured at the iNPUT of the Rod Coleman boards.
Here is the problem. Looking back at the schematic, I can measure 4,8V at the top of the filament resistor and 8.11 V at the cathode of the 201A, and 118 at the plate with respect to ground(0 V) and 30 mV across the resistor R5. If I flip the SW1, it takes the tube rectifier out with a flash. If I connect an interconnect at the output of the preamp to an input of say another amp who's Ground(0V) is connected to chassis ground(IEC ground), it takes the rectifier again. In other words, as soon as I have the chassis ground (IEC) connected to my 0V, it destroys the rectifier!
Also, the 4.85V is the minimum I can get from the Coleman Reg adjustment (using 3.9 ohms R1) and measuring from the Filament Input(+ and -) on the board, I am getting 25Vdc! Why is it too high?
Chassis is aluminum. Heat sinking is to the chassis, umbilical cord is a Neutrik 8 Pole with 1+ and1- as Filament Supply 1, 2+ and 2- as B+ and 0V, 3+ and 3- as HV 0V and IEC safety ground, and 4+ and 4- as Filament Supply 2.
Trying to seek help as I lost three AZ1's already and this project is getting too expensive
Thanks!
Abe
Hard to say without some pictures showing layout and wiring. Specially on ground and earthing.
Some suggestions:
1) Measure the isolation of top FET
2) if top FET is damaged will provide a short or low resistance which may be the issue
3) you should test the gyrator with a dummy load and if you have a variac you can slowly bring HT and find out the issue
3) what anode and mu follower voltages do you have?
Ale
Some suggestions:
1) Measure the isolation of top FET
2) if top FET is damaged will provide a short or low resistance which may be the issue
3) you should test the gyrator with a dummy load and if you have a variac you can slowly bring HT and find out the issue
3) what anode and mu follower voltages do you have?
Ale
IMHO you have some wiring mistake.
What the difference of B+(0V) and HV(0V)?
Why connect preamp chassis to IEC safety ground?
I grounding only PSU box metal chassis, the HV and filament PSUs are independent from earth (floating).
Sometimes connecting preamp metal plate to 0V (aka HV(0V) ) necessary (lesser hum), sometimes not.
My 01A Gen 2 has a 15v DC supply into Rod's regs and a 4.7K sense resistor. Not sure why you're using different values here.
Nothing to do with your problem, but thought I'd mention it.
Hello guys,
Thanks for the responding!
Ketje, no I double-triple check and M1 or M2 does not touch the chassis.
Ale, I realized I do not have replacement rectifiers anymore available(on order) so I cannot do any measurements. I will upload pics later. I was poking around and one thing I did not mentioned was that both L&R channel shares C3 (0.5uF) then connects to the B+. I desoldered it and measure it and I am getting a short. Will that be the culprit? Also, I did test the Gyrator with a 22K 2Watt dummy load before installing it in place to the circuit and it was fine.
Euro21,
Sorry, I made a mistake in my post above. 3+, and 3- is B+ and 0V. 2+ and 2- chassis GND and IEC GND. Let me explain.
For the HV, 0V is the center tap of the transformer secondary. IEC is the safety GND from wall,and chassis ground is the connection of the two chassis. On my 26 preamp PSU, the 0V and IEC ground goes to the umbilical cord to the receptacle connector of the preamp chassis then the IEC GND connects to the middle of the chassis by itself. The OV now is isolated and becomes the GND for the circuit components. The "Chassis ground" is the line that connects the two chassis together.
Filaments 1 and 2 are isolated, after the filter goes to the umbilical cord, then Coleman Reg input, with 1000uF/35V shunt cap, output then to the cathode and the cathode resistor as shown on the schematic. The bottom end of the resistor is then connected to the 0V.
B+ (0V) and HV (0V) are the same right?
Andy,
I emailed Rod and told him I a going to use the regulators for a 01A filament and he replied "...for 01A, simply change R1 to be 3.9 ohm 2W to 5W, wirewound....".
Thanks for the responding!
Ketje, no I double-triple check and M1 or M2 does not touch the chassis.
Ale, I realized I do not have replacement rectifiers anymore available(on order) so I cannot do any measurements. I will upload pics later. I was poking around and one thing I did not mentioned was that both L&R channel shares C3 (0.5uF) then connects to the B+. I desoldered it and measure it and I am getting a short. Will that be the culprit? Also, I did test the Gyrator with a 22K 2Watt dummy load before installing it in place to the circuit and it was fine.
Euro21,
Sorry, I made a mistake in my post above. 3+, and 3- is B+ and 0V. 2+ and 2- chassis GND and IEC GND. Let me explain.
For the HV, 0V is the center tap of the transformer secondary. IEC is the safety GND from wall,and chassis ground is the connection of the two chassis. On my 26 preamp PSU, the 0V and IEC ground goes to the umbilical cord to the receptacle connector of the preamp chassis then the IEC GND connects to the middle of the chassis by itself. The OV now is isolated and becomes the GND for the circuit components. The "Chassis ground" is the line that connects the two chassis together.
Filaments 1 and 2 are isolated, after the filter goes to the umbilical cord, then Coleman Reg input, with 1000uF/35V shunt cap, output then to the cathode and the cathode resistor as shown on the schematic. The bottom end of the resistor is then connected to the 0V.
B+ (0V) and HV (0V) are the same right?
Andy,
I emailed Rod and told him I a going to use the regulators for a 01A filament and he replied "...for 01A, simply change R1 to be 3.9 ohm 2W to 5W, wirewound....".
Some Pics.
IEC Connector to the chassis.: YEL(+) GRN(-) for Filament 1 and BLU(+) GRY(-) Filament 2. ORNG(IEC GND) VIO(Chassis) Ground, RED(B+) BRN(0V).
There is a copper wire than serves as a bus for the OV where the ff is connected from left to right. Input NEG, Output NEG, GNBD for the VOL pot(100k), GND for both Gyrator boards, bottom end of the filament resistor for both channel, then 0V.
0V is not connected to the chassis unless switch S1 is close( left of the Neutrik connector in pic 1.
Notice C3 is not in the pictures as I desoldered it to test it.
Abe
An externally hosted image should be here but it was not working when we last tested it.
IEC Connector to the chassis.: YEL(+) GRN(-) for Filament 1 and BLU(+) GRY(-) Filament 2. ORNG(IEC GND) VIO(Chassis) Ground, RED(B+) BRN(0V).
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
There is a copper wire than serves as a bus for the OV where the ff is connected from left to right. Input NEG, Output NEG, GNBD for the VOL pot(100k), GND for both Gyrator boards, bottom end of the filament resistor for both channel, then 0V.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
0V is not connected to the chassis unless switch S1 is close( left of the Neutrik connector in pic 1.
Notice C3 is not in the pictures as I desoldered it to test it.
Abe
Last edited:
Hi Abe,
A couple of quick checks, with the mains disconnected, and B+ capacitors discharged (measured at 0V):
1. With a DMM set to measure resistance on the 200KΩ scale, measure across the B+ capacitor (red lead to + of the cap). I would expect a reading of more than 100KΩ to show.
2. Disconnect the SUPPLY input to the filament regulator. Measure (again on 200KΩ scale) from the raw dc (+ and -) to the system ground (the 0V of B+ capacitor). This must be open circuit - no resistance reading. Don't power the system if there is a leakage here - this must be cleared.
A couple of quick checks, with the mains disconnected, and B+ capacitors discharged (measured at 0V):
1. With a DMM set to measure resistance on the 200KΩ scale, measure across the B+ capacitor (red lead to + of the cap). I would expect a reading of more than 100KΩ to show.
2. Disconnect the SUPPLY input to the filament regulator. Measure (again on 200KΩ scale) from the raw dc (+ and -) to the system ground (the 0V of B+ capacitor). This must be open circuit - no resistance reading. Don't power the system if there is a leakage here - this must be cleared.
Hi Abe, the mounting of the DN2540 looks a bit messy with that paste around. Is unclear whether you have proper TO-220 isolation and whether the white paste isn't providing any conductivity to the chassis. There is absolutely no need to mount the DN2540 to the chassis on the 01a preamp. You can get away without any heatsink at all.
Is difficult to desolder the devices from PCB if you aren't experienced in doing so. An easy way is to cut the FET pins at the mid point and solder the new pins/legs on top of the old ones, without touching the PCB at all.
My guess is that the DN2540 are fried (quite common) and Rod's test above should help you to confirm this. If resistance is low, you can proceed to replace them.
You haven't placed a drain to source protection zener across the 2sK170, so is likely to be damaged as well.\
Ale
Is difficult to desolder the devices from PCB if you aren't experienced in doing so. An easy way is to cut the FET pins at the mid point and solder the new pins/legs on top of the old ones, without touching the PCB at all.
My guess is that the DN2540 are fried (quite common) and Rod's test above should help you to confirm this. If resistance is low, you can proceed to replace them.
You haven't placed a drain to source protection zener across the 2sK170, so is likely to be damaged as well.\
Ale
Yes, it's certainly true: even if the B+ to chassis test measures open-circuit, it is still possible to get arc-overs from the DN2540's package tab to chassis - usually to the mounting screw. At high voltages (>100V), I try to use a TO220 mounting-clip instead of a screw. The TO220-packaged DN2540 can take >1W without a heatsink, so long as the amp is <50°C inside, so that should be OK.
You can test a DN2540 with a DMM too: on 200Ω range, the D→S reading should be about 11Ω, and G→S should be open circuit (on 200KΩ range).
You can test a DN2540 with a DMM too: on 200Ω range, the D→S reading should be about 11Ω, and G→S should be open circuit (on 200KΩ range).
Hi Rod,
Agree. For this particular circuit, the DN2540 will be running 3mA and even with a voltage drop of 50-100V (VDS), the dissipation is below 300mW.
In my first 01a version (when I didn't have a PCB), I used a TO92 device instead.
Abe
Check as Rod says. If you're lucky the short is between the DN2540 and the chassis. At some point during testing you have connected the SW1 and that may have blown things out. Do you remember when you noticed the issue? If it's just poor isolation, then removing the DN2540 from the chassis will due the trick.
Either way you need to test it carefully before you switch it on again, otherwise you will continue experiencing the same fault again.
Ale
Agree. For this particular circuit, the DN2540 will be running 3mA and even with a voltage drop of 50-100V (VDS), the dissipation is below 300mW.
In my first 01a version (when I didn't have a PCB), I used a TO92 device instead.
Abe
Check as Rod says. If you're lucky the short is between the DN2540 and the chassis. At some point during testing you have connected the SW1 and that may have blown things out. Do you remember when you noticed the issue? If it's just poor isolation, then removing the DN2540 from the chassis will due the trick.
Either way you need to test it carefully before you switch it on again, otherwise you will continue experiencing the same fault again.
Ale
Thank you Ale and Rod!
So far, I have this measurements:
Filament 1: + to 0V =3.36MOhms; - to 0V =10.6M
Filament 2: + to 0V 10.6M; - to 0V 3.4M
GYRATOR
Anode and 0V = 110K
Anode and Chassis = 1.5K
0V to OUT = 110K
B+ and Chassis = 0.2 ohms
GND and chassis = 110k
Yes Ale, you are correct. The first thing when it happened was when I flipped the SW1 switch! And looking at the reading(B+ to chassis) above, it seems to confirm that the short is in the DN2540.
I will take the boards out and do the test both of you posted and replace the parts as necessary. Also, I will get that zener diode installed.
Thank you very much!
So far, I have this measurements:
Filament 1: + to 0V =3.36MOhms; - to 0V =10.6M
Filament 2: + to 0V 10.6M; - to 0V 3.4M
GYRATOR
Anode and 0V = 110K
Anode and Chassis = 1.5K
0V to OUT = 110K
B+ and Chassis = 0.2 ohms
GND and chassis = 110k
Yes Ale, you are correct. The first thing when it happened was when I flipped the SW1 switch! And looking at the reading(B+ to chassis) above, it seems to confirm that the short is in the DN2540.
I will take the boards out and do the test both of you posted and replace the parts as necessary. Also, I will get that zener diode installed.
Thank you very much!
No worries, Abe. You're not the first one blowing off a DN2540 (I can't remember how many I have sacrificed myself during abuse and experimentation). In future, I will recommend to build and test the gyrator board first and then connect it to the circuit. Also if you are bolting any TO-220 device with metal tag on to the chassis, you have to measure the isolation as the tab is generally connected to drain / +HV. Keep us posted!
YEs, it is through the tab of the DN2450 that shorts the B+ to chassis! I took the bolts out from the two boards, lifted the boards, and no more short! I tested the DN2540 just like Rod mentioned. D to S on both is 11 ohms, and G to S is open. But I will replace both when the parts arrive.
Any simple test I can do for the 2sk170?
Thanks!
Thanks Andy, I hope
Any simple test I can do for the 2sk170?
Thanks!
Thanks Andy, I hope
It looks like the fet survived and the short was due to poor isolation on the TO220 tabs (drain). Before replacing the components, remove the boards and test them with a dummy load and see if you can set the output voltage. If the JFET/MOSFET is/are damaged you will get full HT at the dummy load and you won't be able to vary output voltage when adjusting the trimmer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.