[Dr_C]

This thread is intended for digital slotcar enthusiasts who have an interest in repaing faulty digital decoders. They are easy to damage and expensive to replace. Repair is an option for those with a fine tipped soldering iron and basic knowledge of electronics. Next three posts should provide a simple overview of what can be achieved repair-wise.

[Dr_C]

The next three posts provide a simple summary of repairs carried out to a bundle of fourteen faulty Scalextric C8515 digital decoders. 

These were a mix of rev F, rev G and rev H decoders.
Some had previously been upgraded (i.e. MOSFETs) and in addition some had been re-flashed with ICP v3.3.

Faults are typically known to be diode failures, MOSFET failures, and occasionally npn transistor failures - all typically related to excess current and/or voltage spikes.

REPAIR PROCEDURE
===============

The repair procedure was as follows (results will be shared in the next post).

Step 1
=====
Test diodes with a Digital meter (in diode test mode) and replace any that are faulty.

Step 2
=====
Remove both output MOSFETs and replace only the n-channel MOTOR drive MOSFET with a new one. An upgrade replacement AO3434 can be used.

Step 3
=====
If the decoder then functions, i.e. motor is controllable, then replace the p-channel MOSFET with an upgraded replacement AO3401.

Follow-on
=======
If the decoder then functions i.e. motor and brake are controllable the job is done. If not a more advanced diagnostic is required.

ICP
===
Where decoders are flashed with ICP v3.3 an upgrade to ICP v4.0.1 is recommended.

Note: ICP v3.3 was upgraded to ICP v4.0.1 on arrival of the ARC PRO to improve compatibility with this newer powerbase (upgrade features, improved lap counting on ID5 and ID6, more reliable on-track power-up & improved throttle profiles). Neither ICP v3.3 nor ICP v4.0.1 is recommended for use with braided tracks.

[Dr_C]

REPAIR RESULTS
=============

Repair results on the above 14 faulty decoders:

Step 1:
=====
All diodes were found to function correctly - non required replacement.

Step 2:
=====
On removing the two MOSFETs and replacing onIy the n-channel MOSFET, 11 decoders became controllable.

The 3 that didn’t function were investigated and corrected. One had a failed on-board ferrite bead inductor. One had a failed npn transistor (preventing motor turning on). The third had a damaged interconnect track. 

Step 3:
=====
On replacing the p-channel MOSFET again, 11 decoders became controllable in terms of both motor and brake. All three were found to have failed brake enable npn transistors. One failed open circuit so simply the brake would not function. Replacing the npn fixed this problem. The other two had failed as short circuits so the brakes were permanently applied - even when the throttle was applied too. In each the case the npn was replaced and a pair of replacement replacement MOSFETs were fitted (as this brake-on condition can further damage the MOSFETs).

OUTCOMES
=======
14 fully functioning and upgraded C8515 decoders. No write-offs :)

[attachment=32364]

[Dr_C]

CONCLUSIONS
=========

1/ In this sample of faulty C8515 decoders none had failed bridge rectifier diodes.
2/ At least 8 out of 14 (57%) only replacement of the two MOSFETs was necessary to achieve a full repair.
3/ The other main fault was damage to one or other of the two npn transistors that drive/control the MOSFETs
4/ There was one case of a failed on-board ferrite bead inductor and one case of a damaged interconnect track.
5/ The AO3434 and AO3401 as used in the rev H provide a good upgrade path for the earlier rev F and rev G decoder versions.

Hope the above was useful and enjoyable, at least for some ;)

c

[Drifter2]

The first one is the hardest to repair, then it gets easier.  :)

[Dr_C]

If my sample of 14 is typical, it would suggest that simply replacing the two MOSFETs will get more than 50% of failed decoders to run again. Maybe a diagram showing where the MOSFETs are located, and which is which, would help next?

If 20 clubs have 20 failed decoders each, thats’s at least 200 decoders which could be brought back into service with relatively little repair effort. At current decoder prices this corresponds to a saving of £4k or £200 per club. Worth a try?

c

[woodcote]

Definitely worth a try 

I think we have at least 20 failed C8515 and C7005 chips at WHO.

Would be good to see the diagram - and some pics of your repair bench in action.

[Dr_C]

Thanks Andy,

I’ll post a general description of how to remove and replace a three pin SOT-23 device. This is the three terminal surface mount package type we need to remove and replace to swap-out the above MOSFETs.

Not too hard if you can handle a fine tipped soldering iron.

Then, in a later post I will show annotated photos of the three variants of C8515 (rev F, rev G & Rev H) showing how to locate the n-channel MOSFET (Motor PWM) and the p-channel MOSFET (Brake PWM).

c

[woodingboys]

Great work Dr C as always :-) Thank you

[Dr_C]

When Scalextric developed the C8515 rev H decoder they increased the motor/brake drive circuits from circa 1 Amp (as used in rev F and rev G) to 2.8-3.5 Amps. They did this by introducing the AO3434 MOSFET for motor PWM and the AO3401 MOSFET for brake actuation. The AO here refers to the semiconductor manufacturer Alpha-Omega.

In my view, upgrading a standard Scalextric decoder with excessive levels of MOSFET current drive would simply mean that other parts fail first in the event of a short circuit. Even worse if these are difficult to replace parts or cause damage to the circuit board itself. On this basis the AO3434/AO3401 is likely the ideal combination for the Scalextric decoder designs.

Another point worth considering is that Scalextric use MOSFETs which are specified to 30V (Vds_max). Some make the mistake of overlooking this specification parameter and advise a downgrade to 20V. Here we need to keep in mind that due to the use of inductors within the design of the SSD powerbase, a nominally 12V SSD track signal can have voltage spikes up to twice the value (i.e. 24V), so a device with a 20V limit is simply inadequate.

So to cut a long story short the repair advice which follows will use the AO3434 and AO3401 as replacement MOSFETs. These are Scalextric recommended (as per the rev H decoder) and likewise my recommendation for upgrading rev F and rev G decoders).

Photos of the repair procedures to follow.

c