overheating transformer rectifier

Assuming the rectifiers/diodes that were replaced on the advice of the specialist's recommendation are not under rated for their given task, I (and others at a guess) would like to know what was involved in the initial refurbishment two years ago.

I would guess that on some occasions, the replacement of old rectifiers/diodes with their more modern counterparts may require a more efficient (larger?) heat sink and new suitable fan, possibly demonstrated by the fact that an improvement in air flow reduced the likelihood of failure caused by heat.

Were other components "updated"?

Thanks for the comments. I am waiting to get confirmation of the refurbishment but I believe the main change was a new control circuit board and a general cleanup. There is no cooling fan, it relies on passive cooling. We are thinking that this will be the way to go as it would be relatively easy to add some fans. May be you are right and the new diodes need a better heat sink and the overheating before we changed them was due to faulty old diodes.

Twenty year old diodes replaced for more modern ones (without knowing the details of the component) would generally on some occasion require a suited heat sink for that particular diode...the specs for heat dissipation are usually specified inside the manufacturer's datasheet. You can never go wrong with fans.

Would be interesting to see what confirmation you get as far as the refurbishment is concerned. A new for old circuit board would in my opinion not be the cause of your current problem, unless something was incorrectly done there, but very doubtful on that point.

Cheers

Dear,

following may be the causes....

1. Chexk for any modifications done while refurbishment. This may be obstructing the natural cooling. I had experienced even with the layer of varnish poured on it.

2. Check for the quality of diode soldering. End terminal connections. If they are not proper, diode may be heated up , keeping transformer cool.

Even defect in your external connections lead to the similar problem. this nothing but the generation of 3rd harmonics in diode area.

3. Check for the proper installation of new diodes with correct confirmed diode rating with size of heat sink.

Thanks Aurizon for a very comprehensive reply - I have the old diodes so will be able to test them. We are meeting the experts today so should be able to determine their awareness wrt their fundamental orifice!

Hi,

I think you possibly use a diode array of paralleled diodes. If so, there might be a problem with the temperature coefficient (delta Uf/delta T). If they used diodes with a negative tempco there may be a runaway problem of the "best" diode which has the lowest Uf, carries the largest current, lowers the Uf, carries more current, gets hotter, lowers the Uf, +I, +T, -Uf, ... and so on. Overheating is pre programmed then.

Regards Uwe

Could it be that the unit was never fully load tested at 1500 A at 30 V?

At that current , each diode will dissipate 910 watts ( with a forward diode voltage of 0.7V). How many diodes do you have?

From your post you suuggest that with the cover off, the unit is marginally below the temperature limit. My guess would be that the units was never able to meet the full load unless the physical area around the control box was chaged significantly to prevent free air circulation or increase the ambient temperature.

From past experience , I have run into similiar situations where the operators report that a particular item of plant is not achieveing full output only to realize that it never did.

Another situation , with a battery charger, was traced to a root cause of an HVAC unit hot air discharge increasing the general ambient temperature in the area of the battery charger limiting the current. The HVAC unit was installed while the machine was down for a 3 month major overhaul.

Thanks djacob for your comments.

I suspect that you are right that it never achieved the specified output but we have no way of proving it. This unit was made in the UK but has been in Southern Africa for its working life with significantly higher ambient temperatures and lower maintenance & monitoring!

One factor which I only learnt this morning was that the aluminium heatsink and diodes appear to have been laquered during the refurbishment with a 'copper' coloured resin. I am not sure why this was done or if it is normal practice but it strikes me as being one major modification which would affect heat dissipation.

A factor which we see here is where a makeshift or mistaken 'repair procedure' at some in the past becomes the 'norm' from then on so that equipment remains in the wrong condition for ever more. In this case something as simple as replacing the side covers the wrong way up would move two of the air vents to a less effective position but it can be difficult to persuade anyone to change it because that's 'how it has always been'.

Thanks to all for your comments.

Since this unit was made in the UK over 20 years ago, I may not in fact have Shottky diodes. It may use the standard silicon ones.

It may be a good idea to add a ball bearing blower to the heat sink box.

I am not sure what the copper colored varnish was. Ask them what merit such a varnish could have?

I would be very suspicious of ANY resin, paint, etc., applied to a heat sink. Of course, if it is a specialized treatment which only looks like resin, designed to improve heat dissipation, that is different. I have seen some very poor heat sinks in low power situations which were marginally improved by being "painted" black with high temperature paint, but I doubt it would be the same for such high power diodes.

I would very thoroughly and suspiciously question the use of the material on the heat sink, especially if it is any kind of paint, plastic/resin, etc. , and especially, if it is very thick. What material is used for the heat sinks? Is it the same material used in the older design, or perhaps even the same heat sinks?

Thanks Kosimov for your response. The resin appears to be simply a sealant against water or damp. I am waiting for confirmation whether it was applied during the refurbishment or was original. I also doubt that it adds anything to the heat dissipation. It is also the one factor which appears to have changed which could account for the decrease in performance. The heat sink is aluminium and is the original. The problem is that it will prove difficult (and expensive in time and money to correct this)

Will keep you all posted on further developments. Thanks again