Need help with Linn Valhalla repair

Hi Frolic

You may want to replace the BUX85F's as a matter of course too, if one of these is a little 'off' it will take the fuse, bridge and 47R 1W out every time. It may well measure fine at low test voltages when you check it but give problems at the elevated temperatures and voltages of normal Valhalla operation.

I have repaired a few Valhalla's over the years and never had one back - they are quite reliable once sorted. The Valhalla schematic on Vinyl Engine is the complete and correct PCAS022 version, with all the component values.

Hope this helps

Regards
Peter

P.S. The 6A 1000V bridge is a really worthwhile update.
 
Anyone know a current replacement for the BUX84 on my Valhalla?

I've replaced all the caps, the power resistors, the fuse, the bridge rectifier, and put in a thermistor instead of a resistor at r1, and I still blow fuses with my Valhalla. I'm thinking that I need to put in a replacement for the BUX84 transistor at TR2, 3 and 4 on my board (I think it's Q2, 3, and 4 on the schematic).

I can't find any of the part numbers on the schematic or the replacement part numbers, but I was able to find some BUX84 through Newark, but there is a $20 surcharge for getting them from the UK. Is there are more readily available replacement for that transistor that someone else has used? I could spring for the $20, but if I can just use something that digikey, mouser or newark stocks then it would be faster and cheaper.

Thank you for your time and help.

Best,
Aaron.
 
Having trouble repairing my valhalla, getting worse!

Upon reading these forums and others I decided to try to repair my valhalla after the motor started slowing down. I replaced the 47 uF caps with 500v radials which worked fine. Hooked it up and the motor vibrated with more viguor but could not start the table. I could get the platter going with a push and then the motor would keep it going at speed until I turned off the table. I checked the voltage at the motor windings and got 70 volt with the motor spinning. Checked and lubed the motor. I read about checking the voltage regulators input and output voltage and than bang! I fried TR4. Damn!

Searched high and wide for the oem regulators, found them, the thermistor (sg240) and higher voltage bridge rect.(800v 1.2 amp). Replaced them all- no led is lit. Damn, again!

Checked the voltage at C3: over 300 volts and TR3 shows low voltage. Long story long: should I give up? I would like to keep the table all Linn and not go for a Mose if I can avoid it.

Currently, the led does not light and I have nothing at the motor( 2 v measured at red and blue)

I just measured the vdc across:

TR3-21v, TR4-270v, TR2-300v
C1&C2-160v, C3-300v
bridge rect. is 300v

Any help with this would be appreciated thanks.
 
ok. I will try it in my bad englisch.
Its quite simple. Here we go...

First of all i guess, that you tried to replace the rotten electrolytic capacitors.
By ripping off the 220µF from PCB, the four tiny copper lugs on the upper solderpad to ground-area were cut!
This causes a floating Potential. Drawn by the constant current source it rises up to 300V.

Replace the blown components as follows

1. IC 1 MC3346P. Install Socket and insert CA3086 Replacement
2. D2 = 1N4148
3. Z1 = Zener Diode 7,5V
4. C3 = same Reading, but min. 16 or 25V! ( VDD is about 8..9V or higher!)
5 R26 = 33 OHMs (critical, don't alter)
6 remove blown TR4 & TR3, do not reinstall new ones at this time!
7 make sure, that the -Lead from C3 contacts both sides of PCB!!! (Important, otherwise return to Step 1)
8 make sure that the lower Lead of R26 contacts both sides of PCB!!!
9 Install Switch & Led

10 power up the set and press the button ;-)
11 measure VDD at C3 ( 8..9VDC)
12 disconnect from mains and reinstall TR3 TR4

Motor should turn now
13 isolate the set from Live by using transformer ( There is no mains-isolation and this will damage yor oscilloscope )
14 connect oscilloscope probe at C2 (Ground) and C18 (Output)
15 adjust RV1, that sweep is 5..10V below Clipping Point (occurs first on negativ going sine)

That voltage is around 170..180Vss* and far away from 85V+-1 Mantra!!
Corresponding AC Reading is around 65V
Torque enough with this Voltage, lower noise from Motor & enhanced sound from Deck!*

Your main problem is caused in capacity loss of C 14 (1µF), wich obtaines 100% DC feedback and simultaneously
determines Voltage Amplfication given by Divider R21 /R22

Good Luck!

Volker from cologne

*also empirically investigated by Dr. Jazz, Cologne in the mid 80'

p.S. if you have difficulties to obtain BUX84() from stock, simply salvage an energy saving bulb from scrab.
Better ones have the ideal replacement type BUV46 as pair on board ;-)
 
Do you have a circuit diagram? The Valhalla PCAS 022 diagram (available off the net) is representative of the latest Valhalla's and all the component values are shown. The Valhalla is quite reliable once sorted and will give you many years service. I would recommend that when changing the bridge rectifier, you replace with a 6A 1000V device. This is far more robust and reliable. The three large 47uF caps should be 450V devices as apposed to the 250V devices often used, this improves the reliability too! Don't give up, it's worth doing.

First of all i guess, that you tried to replace the rotten electrolytic capacitors.
By ripping off the 220µF from PCB, the four tiny copper lugs on the upper solderpad to ground-area were cut!
This causes a floating Potential. Drawn by the constant current source it rises up to 300V.

Replace the blown components as follows

1. IC 1 MC3346P. Install Socket and insert CA3086 Replacement
2. D2 = 1N4148
3. Z1 = Zener Diode 7,5V
4. C3 = same Reading, but min. 16 or 25V! ( VDD is about 8..9V or higher!)
5 R26 = 33 OHMs (critical, don't alter)
6 remove blown TR4 & TR3, do not reinstall new ones at this time!
7 make sure, that the -Lead from C3 contacts both sides of PCB!!! (Important, otherwise return to Step 1)
8 make sure that the lower Lead of R26 contacts both sides of PCB!!!
9 Install Switch & Led

10 power up the set and press the button ;-)
11 measure VDD at C3 ( 8..9VDC)
12 disconnect from mains and reinstall TR3 TR4

Motor should turn now
13 isolate the set from Live by using transformer ( There is no mains-isolation and this will damage yor oscilloscope )
14 connect oscilloscope probe at C2 (Ground) and C18 (Output)
15 adjust RV1, that sweep is 5..10V below Clipping Point (occurs first on negativ going sine)

That voltage is around 170..180Vss* and far away from 85V+-1 Mantra!!
Corresponding AC Reading is around 65V
Torque enough with this Voltage, lower noise from Motor & enhanced sound from Deck!*

Your main problem is caused in capacity loss of C 14 (1µF), wich obtaines 100% DC feedback and simultaneously
determines Voltage Amplfication given by Divider R21 /R22

Good Luck!

Volker from cologne

*also empirically investigated by Dr. Jazz, Cologne in the mid 80'

p.S. if you have difficulties to obtain BUX84() from stock, simply salvage an energy saving bulb from scrab.
Better ones have the ideal replacement type BUV46 as pair on board ;-)

Thanks Peter and Volker for these great hints !!

:cheers:
 
I know this is a really old thread, but I have discovered a problem with the Valhalla circuit implementation. It also affects other power supplies based on the Valhalla design, such as the Hercules II.

The problem is poor starting ability due to clipping of the power supply output waveform for the first few seconds after every start button press. This is caused by an inappropriate time constant in the coupling of the last filter stage to the amplifier driver stage.

The PCAS 022 circuit uses a resistor ladder to bias the U2 driver stage, with a feed resistor value of effectively ~220k (R23). The coupling capacitor from the previous stage is 22uF (C15). This gives a time constant of 4.8 seconds. The problem arrises because the DC level from the preceding stage varies when the start button is pressed, and the DC bias point for the output stage takes ~5 seconds to settle down. During this time the output waveform may clip and oscillate.

The fix is to change R23 to 10k and C15 to 1uF, giving a time constant of 0.01 seconds for the DC conditions to stabilise after switch on. That way the motor has full, undistorted voltage right from the button press and produces full torque immediately. Check that the output voltage is still in specification after doing this modification and adjust if necessary.
 
If a lower starting torque is a good thing, there are many ways that starting torque could be controlled progressively. From an electronic engineering perspective, the motor-boating of the Valhalla amplifier circuit at start-up smacks of inept design.

Thank you Johnmath for your very interesting info. I will try this mod someday soon after I remove the Valhalla (DIY fully restored some years back) and test this on the bench with a spare motor. :)
 
The fix is to change R23 to 10k and C15 to 1uF, giving a time constant of 0.01 seconds for the DC conditions to stabilise after switch on.

I did another LP12/Valhalla installation today and used 22k for R23 and 2.2uF for C15. It might be my imagination, but I prefer the performance with these values with a slightly longer time constant (but still 1/100 the original), and will use them for all future Valhalla rebuilds and services I do.

Today's LP 12 was a very early deck with old low-torque motor. After the Valhalla mods and a motor bearing clean and re-lube the platter reaches speed within one revolution with no assistance which is quite satisfactory. I have still to clean and lubricate the main bearing too so it may get even better.
 
Today's LP 12 was a very early deck with old low-torque motor. After the Valhalla mods and a motor bearing clean and re-lube the platter reaches speed within one revolution with no assistance which is quite satisfactory. I have still to clean and lubricate the main bearing too so it may get even better.

I've noticed most early decks have motors with a spring loaded bottom thrust bearing. Due to age, it can wear out, dry out, distorted or worn bearing or gummy grease inside. I remove this and do a mod on it for a smoother running motor. I will also use compressed air to blow out the top bearing sleeve area of any contaminants and lube it with thin motor or machine oil. The bottom bearing mod involves shortening the plastic bearing cap, enlarging the bore and inserting a suitable ball bearing to maintain a tiny but acceptable vertical shaft play. The spring and old bearing is discarded. Just a ball bearing inserted. A tiny bit of good quality MSO2 grease is applied where the ball bearing and shaft meet, then glue it back with super glue or Loctite. The motor should be quite smooth and offer least resistance during start up. I usually test run the motor for a couple of hours before refitting. My deck motor is from 79 and been troublefree and consistent.
I've also found this motor would register different and varying phase voltage difference when different makes or types of 0.22uF phase cap is used. I had tested various types on hand just for fun and to see how much variance. Don't think I can suggest which one is the best. Its like matching resistors or capacitors to suit the application.
 
Last edited:
Thanks for the tips coolmaster. There is a minor problem with the bearing ball approach - the bottom of the motor shaft needs to be flat or convex to provide a point contact for the bearing ball, but in fact the motor shaft has a cup. The original thrust loading mechanism has a point that sits in the cup, but a bearing ball will have an annular contact ring - not ideal. I reduce the spring length to reduce the pressure to the minimum that provides adequate shaft loading and use the original device with a MSO2 paste lubricant. I fond that the bearings need to be stripped of old sticky lubricant with a suitable solvent and I re-lube with a high performance synthetic friction modified shaft oil.

Capacitors of the type used to provide the second phase typically have a +/- tolerance of 10% or worse, meaning a capacitor marked 2.2uF may have an actual value in the range from ~2.0 to 2.4uF. Capacitors need to be selected on test by measuring the actual value of capacitance at mains frequency. Then the match to the motor should be verified by scoping the waveforms and ensuring that the phase-shift of the phase winding is actually 90 degrees. Warning: dangerous voltages are involved.

A capacitor that measures higher voltage compared to another will have a higher capacitance value and consequent induce a lesser phase shift, but that is not necessarily better. The optimum smoothness from the motor occurs when the phase shift is at 90 degrees. The voltage across the phase winding is a function of the reactance of the windings and the capacitance - it can't be used to set the correct capacitance.

Minimising the motor torque variations by correctly setting the phase angle reduces measured wow & flutter, which is measured by analysing frequency variations from playing a 3150 Hz test tone recording.
 
ok. I will try it in my bad englisch.
Its quite simple. Here we go...

First of all i guess, that you tried to replace the rotten electrolytic capacitors.
By ripping off the 220µF from PCB, the four tiny copper lugs on the upper solderpad to ground-area were cut!
This causes a floating Potential. Drawn by the constant current source it rises up to 300V.

Replace the blown components as follows

1. IC 1 MC3346P. Install Socket and insert CA3086 Replacement
2. D2 = 1N4148
3. Z1 = Zener Diode 7,5V
4. C3 = same Reading, but min. 16 or 25V! ( VDD is about 8..9V or higher!)
5 R26 = 33 OHMs (critical, don't alter)
6 remove blown TR4 & TR3, do not reinstall new ones at this time!
7 make sure, that the -Lead from C3 contacts both sides of PCB!!! (Important, otherwise return to Step 1)
8 make sure that the lower Lead of R26 contacts both sides of PCB!!!
9 Install Switch & Led

10 power up the set and press the button ;-)
11 measure VDD at C3 ( 8..9VDC)
12 disconnect from mains and reinstall TR3 TR4

Motor should turn now
13 isolate the set from Live by using transformer ( There is no mains-isolation and this will damage yor oscilloscope )
14 connect oscilloscope probe at C2 (Ground) and C18 (Output)
15 adjust RV1, that sweep is 5..10V below Clipping Point (occurs first on negativ going sine)

That voltage is around 170..180Vss* and far away from 85V+-1 Mantra!!
Corresponding AC Reading is around 65V
Torque enough with this Voltage, lower noise from Motor & enhanced sound from Deck!*

Your main problem is caused in capacity loss of C 14 (1µF), wich obtaines 100% DC feedback and simultaneously
determines Voltage Amplfication given by Divider R21 /R22

Good Luck!

Volker from cologne

*also empirically investigated by Dr. Jazz, Cologne in the mid 80'

p.S. if you have difficulties to obtain BUX84() from stock, simply salvage an energy saving bulb from scrab.
Better ones have the ideal replacement type BUV46 as pair on board ;-)


just a question for you .
on the 1. IC 1 MC3346P. Install Socket and insert CA3086 Replacement .
i can see the MC3346P is ready available . why do you susggest the CA3086 replacement ?
thanks
 
ok. I will try it in my bad englisch.
Its quite simple. Here we go...

First of all i guess, that you tried to replace the rotten electrolytic capacitors.
By ripping off the 220µF from PCB, the four tiny copper lugs on the upper solderpad to ground-area were cut!
This causes a floating Potential. Drawn by the constant current source it rises up to 300V.

Replace the blown components as follows

1. IC 1 MC3346P. Install Socket and insert CA3086 Replacement
2. D2 = 1N4148
3. Z1 = Zener Diode 7,5V
4. C3 = same Reading, but min. 16 or 25V! ( VDD is about 8..9V or higher!)
5 R26 = 33 OHMs (critical, don't alter)
6 remove blown TR4 & TR3, do not reinstall new ones at this time!
7 make sure, that the -Lead from C3 contacts both sides of PCB!!! (Important, otherwise return to Step 1)
8 make sure that the lower Lead of R26 contacts both sides of PCB!!!
9 Install Switch & Led

10 power up the set and press the button ;-)
11 measure VDD at C3 ( 8..9VDC)
12 disconnect from mains and reinstall TR3 TR4

Motor should turn now
13 isolate the set from Live by using transformer ( There is no mains-isolation and this will damage yor oscilloscope )
14 connect oscilloscope probe at C2 (Ground) and C18 (Output)
15 adjust RV1, that sweep is 5..10V below Clipping Point (occurs first on negativ going sine)

That voltage is around 170..180Vss* and far away from 85V+-1 Mantra!!
Corresponding AC Reading is around 65V
Torque enough with this Voltage, lower noise from Motor & enhanced sound from Deck!*

Your main problem is caused in capacity loss of C 14 (1µF), wich obtaines 100% DC feedback and simultaneously
determines Voltage Amplfication given by Divider R21 /R22

Good Luck!

Volker from cologne

*also empirically investigated by Dr. Jazz, Cologne in the mid 80'

p.S. if you have difficulties to obtain BUX84() from stock, simply salvage an energy saving bulb from scrab.
Better ones have the ideal replacement type BUV46 as pair on board ;-)

Hi,

On item 5 R26 = 33 OHMs (critical, don't alter), why? The latest revision of the schematic shows a value of 22 ohms.