Sick Akai GX-635

M

mpbird

New Member
Hello to everyone on the Tape forum. I’m the new guy and I have a ‘sick’ Akai GX-635D.

I’m the original owner but it has seen a lot of storage time over the years.
I’ve looked around on the internet and in this forum and it seems as though I’m suffering from the ‘dead transistor’ syndrome. I’m getting very, very weak playback through a lot of static and am unable to record. When I try to record the VU needles move appropriately but nothing comes out on the tape – I just hear static.

I am not an electronic technician but am otherwise kind of handy. I believe I have the ability to solder new transistors --- which leads to my first question. How the heck do you make the audio board accessible to work on? Disconnecting the cables seems straight forward but there are a lot of wires which are connected directly to posts on the board – i.e. they don’t have connectors on them.

Also I read in another thread about someone who started this project and discovered he only had to replace 2 or 3 transistors. How can I test them to see if they need to be replaced?

Well that’s enough of an ‘introduction’! Look forward to hearing from those who have been down this road before!

Jim
 

Attachments

  • GX-635.jpg
    GX-635.jpg
    90.1 KB · Views: 542
Register to hide this ad
First, try cleaning all the switches and potentiometers with DeoxIT. That was enough to get my GX-1820 going, and it was exhibiting many of the symptoms you describe. I need to pull it apart and replace all of the capacitors and the infamous noisy transistors, but it's working well enough for now.
 
As far as removing the boards, I'll assume there is a way to do the work without disconnecting any of the "permanently" attached wires. In other words, it's usually a better bet to position the board in a manner that you can work on it, not necessarily remove it from the machine. Trying to do this might cause you more grief than it's worth.

Let me say though, that although I've never done any board work on that model, I do own one. I'm also going to be dealing with the noisey transistors sometime in 2010 (I swear!). Don't plan on checking for bad transistors in this model. It's a known issue, so if you're going to do it right, replace them all. They are mostly in matched sets with close tolerance of hFe. The farther they drift apart, the more problems arise. Many multimeters have sockets to check the hFe of a tranny once it's removed from the board, if that's the route you take.

Here is some info on those transistors. I have no interest in this guys sales, and IMHO, his cost way too much. Anyway, he has alot of info on AKAI decks, if you poke around his site.

As Sam said earlier, I'd clean it up with deoxit and then decide how bad your problem really is. Good luck.:thmbsp:
 
Last edited:
Heh...

attachment.php





That looks familiar....:thmbsp:



CIMG7829.jpg




The pix look identical, so can I safely assume you have the "D" model, i.e., the non Dolby/"DB" model?

I'd spend some time and effort in hunting down a service manual. One of the AK members PM'ed me and has offered to provide me with one, but I've not heard back since his initial offer.

Tell you what, PM me and we can talk about splitting the cost of one if I can't find one otherwise.
 
Here's a link that may help you with the transistor issue.
I'm getting ready to do mine and just added it today..

http://obsoletemedia.com/geeklog/forum/viewtopic.php?forum=2&showtopic=826&highlight=2sc458lg

Here's the pertinent part..

"You need to to replace the 2SC458LG transistors in the audio amplifier. These transistors usually blow after 20-30 years. The GX-635D has about 16 to 18 of them; the GX-620, GX-625 and GX-255 only 12.
You can replace them by 2SC1815 transistors without changing the pin layout.
I have done this many times before.
When these 2SC458LG turn bad you usually hear a big hiss in one or both channels during recording and/or playback.
"
 
Interesting..
I really don't know squat about these transistor swaps, but I found this thread linked below and posted in it. Hopefully, we'll get a good answer.

Fred and EW are pretty sharp and should give us the straight poop, but I'm going to guess either the 1815s or 2240s are better than what's in there now...

And the first answer is back...

ecluser said:
The 2SC1815L, which is a low noise version of the 2SC1815, is close to the 2SC2240 in regards to noise performances.

Any one of them, including the regular 2SC1815, is better than the 2SC458LG. Try the 2SC1815s you have.
Click to expand...
 
Last edited:
Thanks

OK, thanks for the info thus far. First course of action will be to try the Deoxit treatment - makes sense rule out the easy fixes first. :thmbsp:

I looked over the deck again this evening - I just don't know about trying to do the transistor thing. I'd hate to ruin it -- on the other hand it isn't doing me a bit of good like it is...... :scratch2:

I'll sleep on it for now.

Thanks again everyone.

Jim
 
I'll post after I do mine. It's been sitting for about a year and I need to get to it.
Some recent work recapping some other PC boards and tube amps has given me the confidence to jump in.

If you haven't soldered much, I'd mess around with some old board and get the hang popping old caps in and out. Better to mess up a hunk of junk than something that matters :yes: BTW, I used some desoldering braid and it works pretty good at cleaning out all the old solder so you can pull the leads out.
 
mpbird said:
OK, thanks for the info thus far. First course of action will be to try the Deoxit treatment - makes sense rule out the easy fixes first. :thmbsp:

I looked over the deck again this evening - I just don't know about trying to do the transistor thing. I'd hate to ruin it -- on the other hand it isn't doing me a bit of good like it is...... :scratch2:

I'll sleep on it for now.

Thanks again everyone.

Jim
Click to expand...

I don't know about your machine, but my 1820 has a slider switch buried in the headblock that could only be seen with the cover off, looking up into the assembly from below. That little bastard was the root of all of my channel dropping troubles, and it took me most of a week to find it.
 
S.M. on the way....

...All 70 pages of it. I'll see if I can get it scanned and converted to pdf format to allow adding it ot the AK database....:thmbsp:
 
Well, I just soldered in 10 new ones and now get no record at all and playback on only one channel. Not real happy about it. I looked at the joints with my loupe and they don't look to be touching, so I dunno.

I did find two types of 458s in there. TR1-3 were the round with one flat side, but backwards legs. TR4 was also a set, but they were the the rectangular ones with the slanted edge. TR5 and 6 were also 458s, but my schematic from HiFi engine said they should be 945s, but since they were 458s I pulled them and subbed in the 1815s too.

Similarily, TR7-9 were supposed to be 458s, but were 945s, so I left them alone.

It took about 2 hours as I missed the higher numbered ones and had to go back in. But all were accessible once the bottom was removed although I had to use my hemostat to get a couple in.

Here's a couple pics. The first is the bottom of the board with the legs of the three sets of TR 1-3 sticking up (two on the rigt have already been trimmed off; you can see the untrimmed ones to the left)

attachment.php



And the transistors. The new 1815s on the top, the backwards round 458s to the right and one of each side of the rectangular 458s on the center and left.

attachment.php


Not sure what the next step is. I might reinstall the old ones or just review the joints and resolder, or maybe even buy some new transistors. Is is possible I cooked these? U just soldered them in w/o any heat clamps and I've heard than some of these guys need a heat sink attached to protect them?? this is my first shot at replacing transistors, but it seemed pretty straight forward, or so I thought....
 

Attachments

  • tranny.jpg
    tranny.jpg
    18.4 KB · Views: 487
  • tranny1.jpg
    tranny1.jpg
    84 KB · Views: 494
Damn!!
I tested some of the transistors this AM and it appears I misread the .pdf as a top not bottom view and installed them all backwards..
I'll report back after I swap them all around

Yep, works a lot better with them soldered in right :)

I still have a weak channel on playback, so I'm going to revisit those connections and clean the heads. hopefully, we'll be all done shortly :D
 
Last edited:
The channel still has playback issues in both directions, but the record function works great, which was the initial issue, so that problem is solved.

I need to study the schematic better to determine which of these guys are in the playback loop and which are in the left channel to focus my work now.
 
Andyman said:
The channel still has playback issues in both directions, but the record function works great, which was the initial issue, so that problem is solved.

I need to study the schematic better to determine which of these guys are in the playback loop and which are in the left channel to focus my work now.
Click to expand...

I don't know if that machine has one, but double-check that aggravating slider switch in the headblock if it's there. Trying to figure out why my channel levels were all over the place made me nuts several nights.
 
YES!!!! It's finished!!

The playback issue was a bad 458 in the TR7 spot, which was the hardest to get at. It is behind the for/rev bar and under the timer mechanicals. You have to remove the solenoid bar and pull out the timer and it's still a bear to access. Thank God for curved hemostats!!

I strongly advise downloading the schematic from HiFi Engine if you wish to replace these transistors; it was of enormous help. The scans are huge too, but you can go to 25% to locate something and then zoom in once it's located.
 
Sam Cogley said:
I don't know if that machine has one, but double-check that aggravating slider switch in the headblock if it's there. Trying to figure out why my channel levels were all over the place made me nuts several nights.
Click to expand...

No it was a bad transistor. TR7 read 66 and the others all were over 240 hFe on my tester; the TR8 paired with it was 308. I decided to test and match the transistors earlier today and that's how I figured out that I installed them all backwards yesterday. Most of the 458 tested OK; I think only 3 or 4 were bad; I know for sure two were, but I swapped them all out for 1815s and am good to go for quite a while.

All in all, not a bad adventure. I did lift two traces, but was able to catch edges and extend the leads to other joints nearby, so it wasn'r much of an issue.

I'd do it again, especially now I know where to look and what to look for.

I still don't know why there's a couple 945s in spots the schematic calls for 458s thought :scratch2:, Especially these 7 and 8s, 458s on the left channel, 945s on the right????
 
Andyman said:
No it was a bad transistor. TR7 read 66 and the others all were over 240 hFe on my tester; the TR8 paired with it was 308. I decided to test and match the transistors earlier today and that's how I figured out that I installed them all backwards yesterday. Most of the 458 tested OK; I think only 3 or 4 were bad; I know for sure two were, but I swapped them all out for 1815s and am good to go for quite a while.

All in all, not a bad adventure. I did lift two traces, but was able to catch edges and extend the leads to other joints nearby, so it wasn'r much of an issue.

I'd do it again, especially now I know where to look and what to look for.

I still don't know why there's a couple 945s in spots the schematic calls for 458s thought :scratch2:, Especially these 7 and 8s, 458s on the left channel, 945s on the right????
Click to expand...

Excellent news! Not sure why they used different transistors, it could have been a running change or they might have been equivalent parts.
 
Congratulations - allways there are a thing hidden i hard to get places ...
well - they do assemble circuit boards outside the box for a reason :)
 
You must log in or register to reply here.
Back
Top Bottom