Rackmaster 260 wrong dim curve

[Stratzzi]

Dear Zero88 Forum,

i need some help for my Rackmaster 260 dimmers.
It seems that some Channels are just starting at 50% to dim up and then they go really fast to full output. 
I attached a video that shows the diffcult channels, the both in the mid are dimming perfectly, the others not.
Maybe someone could help me.

Thanks
Stefan

[KWR88]

Hi Stefan,

that looks like possible preheat issue.

if the pot is adjusted too far either way it casue the ramp for the channel control to distort.

it can either leave channels on at a level, or as you have the output don't come on until you get higher up the curve.

to set the preheat we would normally put a 10% control signal in to all the channels then adjust the pot until the lamps are starting to glow then take the signal to 0 and the outputs should then be no longer visable.

 

 

[iank99]

Has that setup worked before?

The fader control looks like it may be of chinese origin & sometimes the DMX they output isn't 'standard' & can cause erratic behaviour. To check the dimmer properly you need a small analogue desk of some flavour i.e a Level 6 or Level 12.

[Stratzzi]

Thanks for your answers. 

I checked the preheat, but when i turned it up, it dims up like it should be. All pots are at zero.

This setup had worked before! I measured the analog voltage and its the same on all 6 channels. So it couldn't be the multiplexer or dmx controller. I got these problems at another setup too (grandMA2 and other multiplexer). 

Any other ideas?

[KWR88]

If the preheat are set correcty then the next thing to look at would be the supply rail or the vref tail not being correct.

If the v-ref rail is out this will cause issue with the ramp as well. given the age of the unti you may find the capacitors on the supply rail have dried out.

these are the 470uF electalitic next to the transformer.

[Sape]

Just read this topic with great interest. As a volunteer technician in our theater we have a similar problem. If we control 4 identical lights simultaneously, their intensity differs considerably at partial load. We have 6 Rackmasters 260 connected to a Showtec Multi-exchanger and a number of halogen lights ranging from 500 to 2000W per channel.  I haven´t tried to adjust the preheat, that will definitely be a next step. The reason of my response: last week we analyzed the output with an oscilloscope and found weird curves. At 50% DMX signal many channels had no output at all. We hope/expect to improve this with preheat adjustment.

At 75% we actually saw two different patterns: one decent,typical, leading edge dimmer curve, and another one jumping up and down to high peaks with a peak-to-peak voltage of 915V, considerably higher than about 615V from the power supply. The 100% curve was more or less as expected.

My question: what could be the reason for this different behavior @75%? Could it be dried up capacitors mentioned in the latest response? A few comments: In all cases we used the same 1000W lamp. For safety reasons, the oscilloscope signal was connected via a voltage divider consisting of 3x 10k resistors so you have to multiply the voltages in attached images by 3. We haven´t opened a unit yet, that will also be a next step. Your reply is highly appreciated.

[kgallen]

Interesting. I have no answer other than cogitation wondering if this is an issue on the control side - in which case probably this would probably manifest on pairs of channels (1/2, 3/4, 5/6), or the output inductor which controls the rate of current change and hence the rise and fall times. However since they are "simply" a coil of wire, are very reliable, don't deteriorate in value over time like caps do (yes, the enamel could deteriorate), and would manifest on one channel, not across a range of channels.

It's with questions like this that we'll really miss Keith and Ian... I hope you get an answer or with further debug identify what the issue is. I'm interested in the answer either way.

[Sape]

Dear kgallen,

Thanks for your reply. Reading your comments, I checked my measurement notes. Due to time constraints we could not test all 36 channels. On one unit, signals from 1/2 were nice, 3/4 and 5/6 were weird. However on two other units, 3 was nice while 4 was weird and 5 was nice while 6 was weird. So these results do not clearly point the finger to the control side. From what I learned about dimmers is that the triac cuts away a part of the sinus. The remainder should still match more or less the original AC sinus. And that is clearly not the case. Just found a circuit diagram for Rackmaster 260/660. Needs a little closer investigation. It'll take a while, but I'll certainly come back on this topic.

[kgallen]

10 hours ago, Sape said:

So these results do not clearly point the finger to the control side. From what I learned about dimmers is that the triac cuts away a part of the sinus.

The triac will trigger part way through the sine wave so effectively the first part of the rise is cut off. The output inductor controls the rise time of this potentially sharp switch-on. This assumes the triggering circuit, and the diac in-line with the triac gate are functioning correctly. So I wouldn't rule out the control side just yet, particularly with so many channels looking bad. If a one-off channel was bad then I'd immediately suspect the diac, triac or inductor. Multiple channels looking bad in similar ways points more likely to a common problem rather than faults of the same type in the discrete components of multiple independent output stages.

Probing this is dangerous because of the mains voltages, but where competence and safety allows I would take a look at the inputs to the diacs on each channel and compare good and bad channels. Then I would work backwards to the opto-isolator and then onto the low voltage side of the control circuit.

Unless you think there are several different faults you don't need to debug on all 36 channels, just take a Rackmaster which has a good and a bad channel and compare step by step back from the diac.

If you can upload the relevant page(s) of the schematic here that would be useful to give more specific guidance.

Let us know.

Regards, Kevin

[Sape]

Hello Kevin,

Thanks for reply again. I included the circuit diagram which I found in a Rackmaster 660 service manual, but interestingly this print also lists model 260. However, AFAIK the 260 cannot run on a 0-4.3 mA input signal, so may I assume that components related to this input are left out?

The inductor you referred to is probably shown as filter in this diagram. Correct? Studying simple household dimmer circuits I do not find these coils.

In the diagram I cannot find an adjustable component related to the pre-heat function. Am I overlooking something?

Next Saturday we will take one or more unit apart and do some measurement. I suggest checking the multi-exchanger's 0-10V output as a starting point, the TP test point, the point between IC2 and C8 (at least the diagram shows what to expect here), the diac's LED, its input and output. 

It could well be that the 6 units are too tight in the rack so that components suffered from the heat. I also suspect the units haven't been opened since the start of their operation, so we expect to find a lot of dust. I let you know the outcome.

Sape

zero88_rackmaster_schema.pdf

[kgallen]

Hi!

Thanks for the schematic. Your plan seems appropriate.

Yes Filter 1 and Filter 2 should be large (maybe 2" diameter or more) wire-wound inductors (you should find small versions inside a domestic thyristor dimmer).

The diac I mention is part of the opto-isolator, IC4, IC5. They are probably socketed.

Given these are two-channel modules I would trace back on them comparing channel-for-channel as you go. Also confirm the indicated waveform at the TP and the negative pulse output on pin 1 of IC2 as you propose.

If both of the channels of an output board are suspected then check same circuit points on a known good board.

This should be relatively easy to debug by doing channel-by-channel compare and identify if the issue is in the common control section IC1,2,3 and associated passives especially capacitors and diodes, or on the output side, IC4,5, D15-18.

The LEDs L1,2,3 should also give you some info, so again compare channel to channel.

Preheat is R11. I compared with a Betapack 2 schematic and it's largely an identical circuit.

Also make sure switches like S3, S4 aren't incorrectly set or oxidised. They look like channel test buttons. I'm not sure what S2 is for - it looks like a jumper from the PCB layout. Check these are all in the same position, especially on channels that are showing different behaviour.

I'm not familiar with the 0-4.3mA current drive signal, I've not come across this before. Probably R28,29,30 are not fitted if it's not supported. IC2 is there for other circuit uses (it's a quad op-amp, as is IC3).

Kevin

ps I found a manual here if of any use: https://www.theatrecrafts.com/archive/documents/Rack-660-Manual.pdf

[Sape]

Hello Kevin,

Last Saturday we opened one unit that had dimmers with expected curves as well as weird curves and did some testing on these. We first did the adjustment of the preheat. They were all off by far, so once corrected, the faders gave a much better linear control of the intensity. The weird oscilloscope images remained the same at partial loads.

The unit was connected to a single phase 230V which makes sense as the lamps are also 230V. The whole rack containing 6 units is fed with 3 phases + neutral, so my best guess is that each 2 units share a phase. The frequency settings were OK @ 50Hz. The big capacitors looked fine by appearance, no swollen end caps that indicate a problem.

Unfortunately the PCB layout differs considerably from the circuit diagram I posted earlier. Doing checks on the board is not easy as there are many thick wires going across that limit access to the components. Nevertheless I was able to pick up a signal from the LHS opamp at pin #1 (AFAIK) that resembled pretty much the signal shown in the diagram at the test point. We were not able to find the other signal shown on the diagram, so we skipped that. We then checked the output of triacs. These were all absolutely fine for 25, 50, 75 and 100% DMX input! But still, at the socket, near the lamp, the signals were still weird. I must admit we used a 10 m hedge cutter cord to floor level in order to connect the oscilloscope.

Our preliminary conclusion is that the wiring to the grid can cause the problem. In the grid, we have 6 aluminum tubes containing 8 individual sockets each that are wired with long multi-core cables to big Harting connectors in the back of the dimmer rack. Question to you: Are there any recommendations for the wiring, like twisted pairs or shielded? Anyway, the controlability has improved a lot thanks to the preheat adjustment, so we leave it for the moment.

There was one other issue we faced: on the third dimmer set of another unit we were not able to make a synced preheat adjustment for both channels. This was both visible at the control LED's as well as at the lamps themselves. Do you have any suggestion to solve this? For the time being we discontinued the use of this channel.

[kgallen]

Hi @Sape

Just to quickly answer a couple of your questions:

18 hours ago, Sape said:

Question to you: Are there any recommendations for the wiring, like twisted pairs or shielded?

I believe you are talking about the distribution of dimmed mains power to a tungsten lantern hang - i.e. cables from the dimmer outputs to (say) 15A sockets on your internally wired bars. I which case I would expect this to be something like H05 or better H07 TRS (toughened rubber sheathed) 3-core cables with 1.5mm^2 or 2.5mm^2 cores. No, this doesn't need to be twisted pairs or shielded.

If you have any DMX distribution - and I don't think you do, other than lighting desk to demux (or dimmers), then that should be DMX rated cable of 110 ohms characteristic impedance, twisted pairs and shielded (note: sometimes Cat5 ethernet cable is used in fixed installs). Any analogue control distribution (from demux to dimmers) would be 6, 7 or 8 core signal cable with a braid or shield.

 

18 hours ago, Sape said:

There was one other issue we faced: on the third dimmer set of another unit we were not able to make a synced preheat adjustment for both channels. This was both visible at the control LED's as well as at the lamps themselves. Do you have any suggestion to solve this? For the time being we discontinued the use of this channel.

This does sound like a control circuit issue within the dimmer, possibly the channel opto has failed or is deteriorated.

I suggest we request to @Edward Z88 for a copy of the correct set of schematics for your units. Maybe you could share on here some details - exact model number and any PCB numbers and PCB version numbers you can identify from your units. Then we can attempt to debug the circuits. If Edward is not happy to share them with you then maybe he'll share with me and I can try and point you in the appropriate directions to debug. Probably however you should be able to identify the opto isolators on the different units and check around them (beware mains voltages on the output side!!!) from the information you have and translating this accross to the newer/older units. Let me know if you want help interpreting the schematics you have.

Just to confirm if there is any doubt, I'm not a Zero88 employee or service tech - I'm "just a bloke on the 'net"...

Regards, Kevin

 

 

[Edward Z88]

Hi @Sape & @kgallen

1 hour ago, kgallen said:

I suggest we request to @Edward Z88 for a copy of the correct set of schematics for your units.

I have emailed you both a link to the Rackmaster schematic.

[kgallen]

Hi @Sape

Did you get the schematics?

Some of your questions - references to the "new" schematics/PCB layout sent by Edward.

On 7/8/2024 at 7:06 PM, Sape said:

The frequency settings were OK @ 50Hz. The big capacitors looked fine by appearance, no swollen end caps that indicate a problem.

To confirm, for 50Hz UK, K3 is fitted, K2 is not fitted.

 

On 7/8/2024 at 7:06 PM, Sape said:

Doing checks on the board is not easy as there are many thick wires going across that limit access to the components. Nevertheless I was able to pick up a signal from the LHS opamp at pin #1 (AFAIK) that resembled pretty much the signal shown in the diagram at the test point. We were not able to find the other signal shown on the diagram, so we skipped that.

The ramp signal is present on IC1 pins 1,5,9,13 if you wanted to check that now you have updated service information.

 

On 7/8/2024 at 7:06 PM, Sape said:

But still, at the socket, near the lamp, the signals were still weird. I must admit we used a 10 m hedge cutter cord to floor level in order to connect the oscilloscope.

The cable will act as a low pass filter, will pick up a lot of electrical noise and your earth connection on your 'scope channel will be a long way from an ideal reference point for the measurement (a lot of inductance in the ground path). Probably the waveform doesn't look great! But it sounds like you've got things working a lot better after re-calibrating the pre-heat.

 

On 7/8/2024 at 7:06 PM, Sape said:

Our preliminary conclusion is that the wiring to the grid can cause the problem. In the grid, we have 6 aluminum tubes containing 8 individual sockets each that are wired with long multi-core cables to big Harting connectors in the back of the dimmer rack. Question to you: Are there any recommendations for the wiring, like twisted pairs or shielded? Anyway, the controlability has improved a lot thanks to the preheat adjustment, so we leave it for the moment.

The dimmer is designed to drive this cabling. The waveform on a 'scope at that point probably looks terrible, but I'm sure a 500W (etc) tungsten lamp will sort that out and be perfectly happy! Cabling as mentioned in previous post.

 

On 7/8/2024 at 7:06 PM, Sape said:

There was one other issue we faced: on the third dimmer set of another unit we were not able to make a synced preheat adjustment for both channels. This was both visible at the control LED's as well as at the lamps themselves. Do you have any suggestion to solve this? For the time being we discontinued the use of this channel.

Sounds like there is an issue with the ramp generator circuit. When you say LEDs above, I'm assuming you mean the output LEDs (L1, L2) that are closest to the test switches. What about the middle LED (L3)? Is that behaving similarly to that LED on other modules? This would help narrow down if the circuit fault is before or after IC1. If L3 is not operating as other modules, then maybe IC2 is at fault. If L3 is ok, but L1/L2 (plus the lamps) are not good, then possibly IC1 is at fault.

If these ICs are in sockets, then replacement should be easy. They wont be expensive. For example:

https://cpc.farnell.com/texas-instruments/lm324n/ic-op-amp-quadruple/dp/SC17193?st=lm324

To diagnose more you would need to use your scope to look around the ramp circuit (VR1 is the pre-heat adjustment pot). On the new schematics this is the circuit lower left if L3 is bad or upper left if L3 is good but L1/L2 bad.

First however, it might be worth moving VR1 back and forwards a few times in case the carbon track is dirty, then re-try your pre-heat calibration.

 

I think at this point you are now the official Rackmaster service agent, as you will know this unit inside and out!!!

 

Regards,

Kevin

 

[Sape]

Kevin, Edward,

Thanks for the support and the schematic. I was a bit of a puzzle to get access to the file, but I managed. Anyway, your help is amazing.

Back at home, I cannot tell you the PCB part number and version. I do remember seeing 1988 printed on them.

Indeed for 50Hz UK, K3 is fitted, K2 is not fitted. Same for the Netherlands 😀

The ramp signal is present on IC1 pins 1,5,9,13 if you wanted to check that now you have updated service information..... Yes, it was pure luck we found this signal on IC1 pin 1. See attached.

The cable will act as a low pass filter, will pick up a lot of electrical noise... My best guess is there are 24-core (or at least 17 cores if a common earth wire is being used) cables from dimmer outlets to each aluminum tube with 8 outlet sockets in the grid, length probably 20 meter each. Also, for load distribution purposes, each cable is connected to different dimmer units, which on their turn can run on different power supply phases. So there is a mix of AC currents going up and down that will influence each other indeed.

Sounds like there is an issue with the ramp generator circuit. When you say LEDs above,..... I meant indeed L1 and L2. For 17 preheat adjustments I was able to find the setting that the theater lights and L1, L2 were at the edge of illumination simultaneously. For number 18 this was not possible.  L3 was always on, on all units. The LM324 are soldered, not easy to replace. Anyway, the schematic provides sufficient inspiration to check various signals simultaneously like IC1 pin 10 & 12, IC1 pin 8  & 14, IC3 pin 2 & IC4 pin2 and Q1 & Q2 collectors. Interesting to see L1 and L2 are now in series with the optocoupler LED's.

Once again, thank you very much for your help. We can figure it out now, after the holidays.

 

[kgallen]

@Sape

26 minutes ago, Sape said:

Same for the Netherlands 😀

Of course! So sorry! 😍