Another SA-9800 dim display issue

jcnash

jcnash

BUSTED!!
Looks like another SA-9800 has bit the dust in terms of having the dim display issue. Just from some general searching on the forums, sounds like a few of these amps have had this issue. The meter remains EXTREMELY dim even when the dim switch is in the bright position.

However, unlike some of the other amps that have had this issue, the dimness in mine is uniform across the entire meter display including the actual power output indicators. It was working perfectly fine till a few days back. Turn it on over the weekend, and I could barely see the meter lights...just like that.

I did some preliminary digging and found this thread where Mark had provided valuable input, and have done some initial analysis.

http://www.audiokarma.org/forums/showthread.php?t=277855

All measurements on the power supply and display boards are absolutely accurate and match exactly whats mentioned in post # 2 of that thread. At a loss for what to do next!? Hopefully its something thats rectifiable and doesn't need new unobtanium parts like the Fluoroscan display itself!

Anyway, Mark and the other gurus here on AK: Would immensely appreciate your input on what else I can test to hopefully narrow down the issue.

Thanks.
 
Also found this PDF posted by Restorer-John here:

http://www.audiokarma.org/forums/attachment.php?attachmentid=201855&d=1269224033

However, upon measuring voltages on the display board, several of these voltages are off/do not match up with the voltages described in that picture. Here are my display board voltages to aid in troubleshooting. All voltages are VDC unless specified.

Pin : Voltage when switch is BRIGHT: Voltage when switch is DIM
1 : 41.7 : 41.7
2 : 42.0 : 42.0
6 : 36.5 : 36.5
10 : 52.4 : 55.8
14 : 1.7VAC : 8.5VAC
15 : 1.7VAC : 8.5VAC

All other pins (labels for input sources) read +19VDC as expected when the appropriate input is selected.

Hoping this helps with the troubleshooting...
 
The filament voltage should not change. The fact that you are seeing up to 8.5v AC on that is not good. You will cook the filament and vaporise off all the barium coating and it will fail.

I would not be operating the display at all, until we locate the problem.

Those voltages on my pdf were for a perfectly functioning unit. There would be variations depending on the market 110/120 220/240 of +/-15% or so each way, depending on the amp.
 
Those pin 14, 15 voltage filament readings are ground referenced?

read the ac voltage across pins 14 to pin 15.

1.7 plus 1.7 is 3.4v which is about right. (3.2vac and 4.3vac on a sx-3900)

There is an ac voltage pin to pin, as well as a DC voltage that is impressed through the transformer's center tap, it is either grounded, or grounded through a 470 ohm resistor (R7, 1/2 watt) and clipped by an 8v zener diode if the voltage drop across the resistor gets too large. Then the current through the display and the r7 resistor supplies a voltage in the dim switch position. In bright the center tap of the transformer is grounded.

The transformer's AC voltage is controlled by R23, a 330 ohm 1/4 watt resistor. IF you are reading 3.4v across the heater, you might want to explore bringing the ac voltage up a small amount by slightly reducing R23.
The easiest way to do that is to parallel a resistor across R23, start at 10k (330 becomes 320) and don't drop much under 1k (330 becomes 250) ohms.

First check that the AC voltage at pin 17 to ground is the same as pin 18 to ground.
 
Last edited:
Realistically, the method Pioneer used to vary the brightness of the VFD flies in the face of ensuring the type of service life these displays would have been capable of if driven correctly.

I've got three dead SA-9800 VFD displays here from repairs. I've been lucky to be able to locate parts units, but it's getting too hard and expensive for the sakes of a badly designed display to bother any more. They should have used a variable duty cycle drive to vary brightness and they'd all be still working.
 
markthefixer said:
Those pin 14, 15 voltage filament readings are ground referenced?

read the ac voltage across pins 14 to pin 15.

1.7 plus 1.7 is 3.4v which is about right. (3.2vac and 4.3vac on a sx-3900)

There is an ac voltage pin to pin, as well as a DC voltage that is impressed through the transformer's center tap, it is either grounded, or grounded through a 470 ohm resistor (R7, 1/2 watt) and clipped by an 8v zener diode if the voltage drop across the resistor gets too large. Then the current through the display and the r7 resistor supplies a voltage in the dim switch position. In bright the center tap of the transformer is grounded.

The transformer's AC voltage is controlled by R23, a 330 ohm 1/4 watt resistor. IF you are reading 3.4v across the heater, you might want to explore bringing the ac voltage up a small amount by slightly reducing R23.
The easiest way to do that is to parallel a resistor across R23, start at 10k (330 becomes 320) and don't drop much under 1k (330 becomes 250) ohms.

First check that the AC voltage at pin 17 to ground is the same as pin 18 to ground.
Click to expand...

My bad...I assumed the voltages on Restorer-John's image were all referenced to ground. Re-measured, and you're right. The voltage between pins 14 and 15 is 3.7-3.8VAC. Looks like that voltage is as per spec.

This might therefore be irrelevant, but I measured voltage at pin 17 and 18 to ground and saw 51.4VAC on both.

Any other recommendations other than to start looking for a donor unit? :sigh:

Thanks Mark!
 
restorer-john said:
Realistically, the method Pioneer used to vary the brightness of the VFD flies in the face of ensuring the type of service life these displays would have been capable of if driven correctly.

I've got three dead SA-9800 VFD displays here from repairs. I've been lucky to be able to locate parts units, but it's getting too hard and expensive for the sakes of a badly designed display to bother any more. They should have used a variable duty cycle drive to vary brightness and they'd all be still working.
Click to expand...

Looks like this is a VERY common problem with this model. Have you seen the same issue on the SA-8800 and 7800, or is it just the added dim/bright switch on the 9800 that seems to be causing all these issues?
 
I'm curious about the display issue because my unit isn't working either... It's currently at a shop, but I'm worried it's not going to be fixed. All that said to ask... will a SA-8800 display work, or I should say, be a drop in replacement?

I know the 9800 part number is AAV-001, haven't found one anywhere, not even a parts unit to use from. However, I did buy one from a 8800, p/n AAV-002. Hoping it's function like a drop in replacement in the event my tech can't get mine working again.
 
ZeusMutation said:
I'm curious about the display issue because my unit isn't working either... It's currently at a shop, but I'm worried it's not going to be fixed. All that said to ask... will a SA-8800 display work, or I should say, be a drop in replacement?

I know the 9800 part number is AAV-001, haven't found one anywhere, not even a parts unit to use from. However, I did buy one from a 8800, p/n AAV-002. Hoping it's function like a drop in replacement in the event my tech can't get mine working again.
Click to expand...

The 8800 display will not work. The 8800 meter only goes upto 80W and the 9800 goes upto 100W. It won't be a drop in replacement for sure.
 
Last edited:
Any other recommendations other than to start looking for a donor unit?
Click to expand...
design a new display is your other option!! of course it will require a bit of ingenuity, know how and will not look the same but if you do a good job it could look alright.
My thoughts are a 4x20 blue back lit LCD character module and of course a mcu to drive it.
Problem with anything like this is, what would someone pay for such an item? and is it worth it for anyone to take the time to design one, knowing the low demand for such a item.
i am afraid that any donor unit will end up the same way as you have now, knowing the design faults. I personally do not like vfd displays because of price, longevity.
 
Last edited:
Would't it be an option just to get the tube re-gassed or vacuumed (im not sure which). Unless the coating has burnt off too?

The examples I have or have seen have a nodule at one end - I guess this is where the tube was originally nipped off and sealed... surely if your going to the extent of building an entirely new display unit, re-gassing etc could be another option?
 
Would't it be an option just to get the tube re-gassed or vacuumed (im not sure which). Unless the coating has burnt off too?
Click to expand...
i do not think so.
reasons being
who is going to do it and for what cost?
and going through all that you find out
"the coating has burnt off"
designing a new display is a lot of work, if you are capable. if i had a sa-9800 or any vfd design, with this problem, this is what i would do. nice blue leds, that will last a life time. the more i think of it, i would like different colour leds for each channel as they did on the sx-x80 series receivers. lots of new display options. the better you are at mechanical design, the more pro it looks. you could get away with no mcu if you use a lm3915 type part, but basically the ckt is a bunch of window comparators.
you could replace the sig level display with analog meters too, look at the ta7318 data sheet. i think the meters in the sx-x80 would fit and work. if you want white meters,look for some from a sansui g-7000/7500 series receiver, i have a g-7500. i always liked analog meters over discrete element types.
 
Last edited:
I doubt the tube is bad. Unless it's gassy. I have a Pioneer DT-400 clock with a "fluroscan" VFT, which has been plugged in and running continuously as a kitchen clock since 1979. That's over 315,000 hours, and it still lights up the kitchen like a night light. I replaced a couple of caps in it years ago, and repaired the relay. But the tube goes on strong.
 
Ed in Tx said:
I doubt the tube is bad. Unless it's gassy. I have a Pioneer DT-400 clock with a "fluroscan" VFT, which has been plugged in and running continuously as a kitchen clock since 1979. That's over 315,000 hours, and it still lights up the kitchen like a night light. I replaced a couple of caps in it years ago, and repaired the relay. But the tube goes on strong.
Click to expand...

Comforting to know that they can keep on going. I don't think I have any that are on their way out yet...

Though if you wanted to keep the original styling etc of the fluroscan, you could just make a metal stencil plate and control a series of blue LED's behind it to mimic the fluroscan. Maybe it could even use the existing circuit to some extent?

The fluroscan look was one of the things that first caught my eye with Pioneer and I always find the look of them pleasing and professional.

Though good ol' needles are nice too! :music:
 
I also considered having made a black stencil and using bar graph drivers, a custom PCB and blue LEDs, but it simply isn't worth the time and effort.
 
Ed in Tx said:
I doubt the tube is bad. Unless it's gassy. I have a Pioneer DT-400 clock with a "fluroscan" VFT, which has been plugged in and running continuously as a kitchen clock since 1979. That's over 315,000 hours, and it still lights up the kitchen like a night light. I replaced a couple of caps in it years ago, and repaired the relay. But the tube goes on strong.
Click to expand...

Thats what Im hoping. Hopefully Markthefixer can work his magic :)
 
The answer is near the end of the first thread linked to, changing the series resistor setup that makes the +68v into (+28 to +32v (not sure, would have to analyze thread)) using a transistor regulator to make the voltage INSENSITIVE to current drawn by the tube.

As the tube ages, more current is drawn, shifting the voltage points in the original circuit. It's NOT designed for the tube's aging over 30 years!!
 
Ah...unfortunate! Ill stay on the lookout for a replacement display...until then, while the display is pretty dim, it still functions and still looks pretty damn good! LOL.

I definitely don't have the know-how to design a replacement LED display and an MCU etc.

Thanks for your help Mark. Much appreciated!

PS: Ive decided to re-place the components on the power supply board of the 9800...Ill revive my old thread for that if I come up with any questions for ya!
 
You must log in or register to reply here.

Similar threads

AlwaysAnalog
SA-9800 Dim Display: Options for Fix
2
Replies
28
Views
8K
redk9258
R
S
SA-9800 very dim Flourescent Display
Replies
3
Views
597
stevieg330
S
ph2113
Repairing SA-9800 #2: Dim display,Dirty Relay & Poor mechanical contacts
2
Replies
23
Views
2K
ph2113
ph2113
Keizer
SA 9800 Fluroscan display question.
2
Replies
22
Views
1K
Keizer
Keizer
SYK3S
Pioneer SA 9800 Weird Display
Replies
7
Views
1K
SYK3S
SYK3S
Back
Top Bottom