Hi all,
Have just joined this forum, i've been a member of other for a few years now and thought its time i add diyAudio to the list.
I've recently acquired a working TA-5650 for all of $10 at a garage sale along with its matching tuner and manuals. I got it just in time as it was very close to blowing up from massive bias issues due to the well known diodes. I am giving the whole unit a rebuild with all new caps, the updates in the service bulletin, new relay, bias and psu adjusting pots, thermal grease etc, the usual.
The one question i have is biasing the amp after its been rebuilt, i know i can power it up without the v-fets installed, check dc ofset, make sure the biasing system is working and adjust the power supply. However what setting should i have the bias trimmers set to when installing the v-fets? Should i set them both to their lowest setting 0mv which is anti clockwise on the old pots?. I have yet to pull the old pots out and see if this is maximum resistance or lowest yet as i don't know which way they are orientated in the circuit. Looks like lower resistance= lower bias on this unit. I have seen other threads stating you start with the bias set at 0mv however when bench testing the unit before tearing to down it briefly hit 0mv on the left channel and it caused the relay to click off until i restored bias to the channel.
Have just joined this forum, i've been a member of other for a few years now and thought its time i add diyAudio to the list.
I've recently acquired a working TA-5650 for all of $10 at a garage sale along with its matching tuner and manuals. I got it just in time as it was very close to blowing up from massive bias issues due to the well known diodes. I am giving the whole unit a rebuild with all new caps, the updates in the service bulletin, new relay, bias and psu adjusting pots, thermal grease etc, the usual.
The one question i have is biasing the amp after its been rebuilt, i know i can power it up without the v-fets installed, check dc ofset, make sure the biasing system is working and adjust the power supply. However what setting should i have the bias trimmers set to when installing the v-fets? Should i set them both to their lowest setting 0mv which is anti clockwise on the old pots?. I have yet to pull the old pots out and see if this is maximum resistance or lowest yet as i don't know which way they are orientated in the circuit. Looks like lower resistance= lower bias on this unit. I have seen other threads stating you start with the bias set at 0mv however when bench testing the unit before tearing to down it briefly hit 0mv on the left channel and it caused the relay to click off until i restored bias to the channel.
Only 10 USD for working one. You are lucky 
. four pairs of VFets.
Do you have service manual? you can down load it from HiFi Engine - Owners and Service Manuals
If you replaced all the small capacitors , double diodes and preset. Measure the ESR of main filter capacitors, if required replace it.You have to check all supply voltages including bias supply. TA4650/TA5650 uses floating voltage doubler circuit for biasing.Add a diode to voltage doubler circuit to make the bias supply more stable.
Before putting the Vfets back measure the source-gate voltage of Vfets (without putting Vfets) and adjust bias preset, it will vary from 10-28 Volts. Set it to highest voltage.if every thing is okay you can put vfets back and adjust bias preset to get required bias. Refer service manual for more details.
If you need any help you can PM me.
. four pairs of VFets.Do you have service manual? you can down load it from HiFi Engine - Owners and Service Manuals
If you replaced all the small capacitors , double diodes and preset. Measure the ESR of main filter capacitors, if required replace it.You have to check all supply voltages including bias supply. TA4650/TA5650 uses floating voltage doubler circuit for biasing.Add a diode to voltage doubler circuit to make the bias supply more stable.
Before putting the Vfets back measure the source-gate voltage of Vfets (without putting Vfets) and adjust bias preset, it will vary from 10-28 Volts. Set it to highest voltage.if every thing is okay you can put vfets back and adjust bias preset to get required bias. Refer service manual for more details.
If you need any help you can PM me.
First thing to do is take out the VFETs - the remains of the thermal compound will have to be replaced anyway. Be careful as they also partly hold the printed circuit board.
Then adjust the bias pots to maximum voltage between S-G of the VFETs (the pins). The adjecent VFET pairs are in parallel (1-2, 3-4, 5-6, 7-8) so you can measure any one that is convenient. Do not return the VFETs until you do a rebuild, the VFETs are far too expensive and difficult to find to risk damaging them. The amp circuit (driver) will work even without them but of course with no or very little load, such as headphones. You can use this to initially check all voltages and DC offset. Then, first mount only one pair per channel just to verify the basic operation (bias adjustment) setting a very low bias (just about when current starts flowing, a few mA). Depending on the rank of VFETs it may take quite a bit of adjustment trimmer travel before current starts to flow! A 10-turn pot is a good idea here.
Also, joshvi gave you very good advice re adding a diode to the voltage multiplier, this was discussed on the forum already in other related threads.
Also, you cannot measure the bias current if the output protection relay is not energized or does not have good contact once it is. A better way to measure bias current for a channel is to measure the voltage between the cases of the 2SJ18 and 2SK60 VFETs in that channel, however, for the same bias current, expect twice the specified voltage. This is because the original method looks at the voltage drop through one of the drain resistors, while the one i propose does this for both drain resistors in series. This method of measurement measures the voltages before the relay so it cannot interfere. Be careful that your probe does not slip because connecting the drain to the heatsink is equal to an output short circuit with the over-current protection disabled!
Then adjust the bias pots to maximum voltage between S-G of the VFETs (the pins). The adjecent VFET pairs are in parallel (1-2, 3-4, 5-6, 7-8) so you can measure any one that is convenient. Do not return the VFETs until you do a rebuild, the VFETs are far too expensive and difficult to find to risk damaging them. The amp circuit (driver) will work even without them but of course with no or very little load, such as headphones. You can use this to initially check all voltages and DC offset. Then, first mount only one pair per channel just to verify the basic operation (bias adjustment) setting a very low bias (just about when current starts flowing, a few mA). Depending on the rank of VFETs it may take quite a bit of adjustment trimmer travel before current starts to flow! A 10-turn pot is a good idea here.
Also, joshvi gave you very good advice re adding a diode to the voltage multiplier, this was discussed on the forum already in other related threads.
Also, you cannot measure the bias current if the output protection relay is not energized or does not have good contact once it is. A better way to measure bias current for a channel is to measure the voltage between the cases of the 2SJ18 and 2SK60 VFETs in that channel, however, for the same bias current, expect twice the specified voltage. This is because the original method looks at the voltage drop through one of the drain resistors, while the one i propose does this for both drain resistors in series. This method of measurement measures the voltages before the relay so it cannot interfere. Be careful that your probe does not slip because connecting the drain to the heatsink is equal to an output short circuit with the over-current protection disabled!
Last edited:
A massive thank you to both of you for your replies. I've been searching for this information for days. @joshvi regarding the filter caps, all electrolytic caps including filter caps are being replaced with Elna silmic II and 80v 10,000 uf nichicons for the filters. Omron my-02-dc24 for the relay, bournes 2.2k multiturn pots for bias and psu, it says 22k for the psu on the schematic but 2.2k on the trimmer in the psu itself, a misprint perhaps? 1n4148 diodes to replace the double diodes, I've got new mica insulators and thermal grease and will be removing the plug on the main amp board and wire wrap soldering all the connections.
@ilimzn When i ajust the bias pots to maximum voltage between S-G, once the vfets are installed does that yeild maximum bias or zero bias?
My plan before all this is to install all the new parts, set the new psu trimmer as close to the resistance of the old one then fire it up without the vfets and set the 20v correctly then proceed with all other tests. Now its just a waiting game for all the parts to get here before i can do anything. Capacitors are on order from parts connexion in Canada, will take about 10 days to get to Aus, the filter caps are on back order from RS components until the 25th so sadly i cant go any further for now.
@ilimzn When i ajust the bias pots to maximum voltage between S-G, once the vfets are installed does that yeild maximum bias or zero bias?
My plan before all this is to install all the new parts, set the new psu trimmer as close to the resistance of the old one then fire it up without the vfets and set the 20v correctly then proceed with all other tests. Now its just a waiting game for all the parts to get here before i can do anything. Capacitors are on order from parts connexion in Canada, will take about 10 days to get to Aus, the filter caps are on back order from RS components until the 25th so sadly i cant go any further for now.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.