Alpine PDX M6 - 43VDC Output

There is limited information on repairing only the Alpine PDX (gen2), so I figured make a thread only for the PDX. Long story short - I found out after getting my JL 13TW5V2 repaired by JL for $230 :mad: that the amp was the reason why the sub failed (melted cone + snapped voice coil), not my occasional listening preference. If you hear a loud turn-on thump while using a PDX, turn your head-unit off immediately to preserve your speakers. The protection circuitry is only to prevent an over-current, over-voltage, and ultimately thermal runaway condition for safety and liability purposes - its not trying to protect your speakers.

My PDX M6 outputs a constant 43 VDC voltage, regardless of whether any signal input is provided to the RCA input. After reading through various other experiences with this amp, and the MRV (or MVR) which uses the same class D switching, I am planning or have already:

1. Done - Exercised the theory of having a loose subsonic switch 'in the middle' of the three state options. With only power/remote connected, the speaker output was still constantly 43VDC, while switching between 30 Hz to 15 Hz to off did nothing. It was set at 15 Hz.

2. Done - bench power the unit at 14.4VDC, then probe the audio driver ICs to check whether they are working properly. I've read through issues which other's have experienced on both amps, and it sounds like the original PWM audio driver ICs can go bad rather easily. I have the M6 Service Manual with all the test point voltages - so I may as well test. Update..PWM driver ICs were ok, all 8 FETs were ok, spot checked some other components on the bottom board and they were ok. I didn’t check anything on the top because that ribbon cable is so short...I’m sure a tech would have a longer ribbon cable for proper testing. Seems like the problem worked itself out once I fed it 14.4VDC.

3. To-Do - Replacing both 98-1036TRPbF PWM audio driver ICs with an IRS20965S, then test to see whether the 43VDC goes away. If it doesn't go away, I may end up just sending it in for repair.

4. Optional - Replacing all 4 Power Supply FETs and all 4 output FETs. If the amp resumes operation after replacing the IC's, I may still go ahead and replace the FETs to refresh the amplifier - hopefully clock in a higher true power rating. The FETS all look good (they aren't brown or cracked), don't have any shorting, but 'fail' a handful of FET tests - caveat is that they are still in-circuit.

If anyone else has any other thoughts or has experienced the same failure, please chime in to provide some assistance. Thank you in advance!
 
Last edited:
Thoroughly tested the M6 board yesterday - within a minute of providing a clean 14.4 VDC source voltage, the speaker output's DC offset which had appeared hard-locked at 42.5 to 43 VDC immediately at receiving power, suddenly went away, and became 0 V. Idle current was approximately 1.5 A. I then proceeded to test the output PWM driver ICs, the output FETs, the power supply FETs, and spot check a couple other ICs on the bottom - the terminal voltages all were pretty close to the call-outs in the Service Manual. Nothing was misbehaving or out of range. There are a handful of call-outs that simply state 'red-led', but I got a voltage (I didn't have the Alpine badge board plugged in).

I then let the amp idle for about 2 hours while powered. The output FETs warmed up gradually to 95 F (thats with the heat-sink blocks attached). Finally tested it with a signal and it was fine - back to normal.

My theory - and thats all this is - perhaps the Alpine's software running on the uCOM gets confused when certain conditions are met which are needed for thermal protection and over-current protection. Once certain values are read by the uCOM (and we don't know what they are), perhaps instead of throttling back the gain, it somehow causes a full saturation the output section. To clear the condition and reset it, simply provide the amp a true 14.4 VDC, under no load? I don't know. I was also pressing firmly on the PWM driver ICs while it was idling and warm - no reaction, so no cold solder joints.

Plays music fine from vehicle power which is only 12.5 VDC, and 13.8 VDC with the engine on, and when starting from 35 F - guess I find out tomorrow morning what it does when starting from 20-25F.
 
It’s a common fire hazard problem then- a handful of guys have wrote in the Amazon reviews they have the same exact failure mode. Mine was manufactured 2015, only months before I purchased it...this is a 2010 design, so the defects and bugs should have been worked out by then. It’s got to be a fundamental design flaw. Or could this be a leaky set of FETs?

I used it for a few hours today and it was fine- although I didn’t push it at all.
 
Many amps have DC offset protection. Others rely on the fact that the servo/feedback circuit will drive the outputs as hard as possible to counter the offset (and bring it to 0v). When something fails (outputs short, etc...). This causes extreme current through the outputs and the over-current protection engages or the fuse blows.

It's possible that there are leaky outputs. It's odd that the amp got up to 95 degrees at idle (unless ambient was very near 95 degrees).

When the amp initially had rail then it went back to 0v, was a speaker connected?
 
Thank you Perry for the technical info - so do you think the FETs being warm under idle is something to be concerned with ? What could be causing them to be become warm at idle?

Listened to the amp and sub today - it sounds fine, although I'm keeping an ear out for that hard turn-on thump now.

I have all my voltage measurements written down for the two PWM driver ICs and all the FETs, and post them tonight in case anything might seem like its out of range. There were two pinouts on the IC's that should have been -50.X VDC, which were -40.X VDC...I don't remember off the top of my head.

In my researching trying to find the Toshiba uCOM datasheet (it would be nice to know how this works and possibly dump its firmware) I found the Alpine PDX F4 service manual - so I have that.
 
I'm just curious - the same microcontroller is used in the lower cost Alpine amps (recent stuff from 2010's) as well (according to some repair manuals I found). I want to see what the DC Offset inputs, iLimit inputs, ect. are reading into on that chip. Maybe there is programming/extracting info as well. I also want to see how the uCom reset input works (what is the uCOM looking for from the separate uCom reset chip arrangement).
 
The PDX and refurbed sub have been fine, but the 43VDC problem came back temporarily this morning - I think its influenced by minor cold temperature shock: 50 F weather yesterday, which rapidly swung to 25 F overnight and stayed there (I park outside), the truck and amp are cold soaked at 25 F. As soon as the amp turned on cold with no music signal, the sub pushed out. Shut off the radio, and I could hear the sub return to its rest. After 15 min of driving, and flipping the radio on and off, eventually the amp didn't saturate the output, and I was able to listen to the stereo at a reasonable 'morning' level.


The amp has an intermittent issue which will be very difficult to chase. I suppose I could stick the amp in the freezer (in a ziplock)) for overnight, then bench test it. Keep in mind that the board had no glaring or visual clues - it still looked new inside. Should I just swap out the output FETs as a starting point and see if the problem goes away?
 
Dusting off this old issue, which was never corrected. I was able to finally shake the failure permanently by listening to some bass heavy music (in hopes of thermally shocking the degraded components) loudly while the truck (and amp) was initially cold-soaked at 10F to 15F. This DC-only output issue doesn't seem to be temporary any more. I had been dealing with randomly not being able to listen to music, books, or anything for periods of time because this amp would simply output DC, and I'd have to shut off the radio to prevent speaker damage.

I had it on the bench yesterday - I could hear a rapid ticking/clicking noise coming from the board only while the speaker-plug was connected (always make sure the speaker-plug is connected while probing!). Two of four Output FETs were saturated to +50 VDC and/or -50 VDC on multiple legs of the same FET, and which doesn't align with the Service Manual voltages listed. (I didn't write down any of the voltages or which legs as I time-crunched and testing with the board upside down). Power FETs were ok.

All new Output FETs are going in today. Any benefit to changing out the Power FETs while I'm in the amp?

I'll have a new thread for the Alpine PDX F4 (for fluctuating volumes on channels) going up once this M6 is fixed.
 
I'm confident that there is an issue with output FETs since one of the legs was supposed to measure as 0 V, but it actually was measuring + or - 50 VDC, which was the same as the neighboring leg. Two of the four output FETs did not have the correct terminal voltages for all three of the FET legs.

I'll leave the Power FETS alone then.
 
Last edited: