Replacement/alternative power supply for LPL Saunder 4500 ii (aka 7451) enlarger

[brianssparetime]

I recently obtained an LPL / Saunders 4500ii (aka 7451) enlarger with VCCE head, but the previous owner lost the power supply brick. The 4500 ii uses a 200W 82V RMS AC bulb.

I’d like to make a replacement power supply. While there are a few listings out there for replacements, they are all very expensive ($500-$1000). I’ve previously designed and built my own arduino-based darkroom timer (including making a custom PCB), but I have very little knowledge/experience with this kind of AC circuitry.

Questions: Does anyone have pictures of the innards of a 4500ii 200W 82V power supply? I would also appreciate any advice on the 5 approaches laid out below - either on the relative merits, or on the implementation details.

Based on my research, I have a few options:

1. Use a simple AC dimmer and a dummy 200W load to bring my 120V line voltage down to the right level
2. Use a 3M overhead projector (OHP) power supply, or build one based on this design. The 3M projectors use bulb with the same voltage requirements, though slightly higher wattage.
3. Obtain a generic variable regulated/stablized power supply, which can be adjusted to the right voltage.
4. Replace the bulb with a 24V bulb, and use a more commonly available 24v power supply.
5. Design / build a regulated stablized power supply.

As more background, “82V RMS is 1/2 wave rectified 117V. In cheaper, unregulated, heads there is just a diode to drop the voltage by cutting out the negative (or positive) AC cycles. When used in regulated or closed-loop heads the power to the lamp is controlled by a triac that fires at a nominal 90 degrees. At 90 degrees on the AC waveform controlling the average voltage does a reasonable job of controlling the RMS voltage. The voltage is actually 82.7V - but the diode (or triac) takes away 0.7 volt, leaving 82V to go to the lamp.” (https://www.photrio.com/forum/threads/lpl-4500-dying-power-supply.37016/)


Option 1 (dimmer) seems the easiest. There are reports of this working (https://www.photrio.com/forum/threa...roic-power-supply-options.166434/post-2166293), though it would not be stabilized or regulated, it’s pretty inefficient, and I have some concerns about heat dissipation.

Option 2 (OHP): I can source a 3M projector for $100, or it appears the power circuitry is very simple, consisting of “a P1000J silicon rectifier (600PVI, 10A) in series with the lamp and line and a P6KE400CA transient voltage supressor (400V, 600W) in parallel with the bulb.” (https://www.diyaudio.com/community/threads/115-volt-to-82volt.28924/). This would not address fluctuations in the input line voltage, and I’ve read that there may also be issues with spikes when it first turns on.

Option 3 (generic): I haven’t been able to find a suitable generic power supply, but maybe I’m searching the wrong terms. I believe an SCR type should do this, but lab power supplies tend to be very expensive as well. If one is available at a good price, I would be very happy to go this route.

Option 4 (24v 200w bulb): This was suggested here (). I believe some Besseler and Durst enlargers use 24v supplies. If anyone has recommendations for particular 200W bulb/PS combinations, I’d love to know.

Option 5 (full DIY): The longer quote I pasted above as background outlines the use of a triac firing 90 degrees from the AC waveform as the basis for more sophisticated power supplies. Following up on that, I found this page (https://www.homemade-circuits.com/how-triac-phase-control-works/) (see Figure 4) that shows a power supply circuit diagram using a zener to stabilize it against input fluctuations, and a triac to produce the output wave. Interestingly, the article comments that the “set up helps to sustain virtually a constant output regardless of any significant variations in mains AC input voltage. It is regularly found in photographic and other applications where a highly stable and fixed level of light becomes essential.”

However, I suspect that this is one of those textbook circuits simplified to illustrate the principle instead of optimized for real-world usage.

This article (https://eepower.com/technical-articles/alternating-current-ac-load-control-with-triacs/#) (see figure 3) has a similar circuit it describes as a “UJT trigger circuit” for the triac, but I don’t really understand enough of the commentary around it to know if it’s appropriate.

Beyond figuring out what values of components I’d need to output 82v rms (what resistor/cap/zener ratings, what kind of pulse transformer, and UJT), I’m curious if there are other bits I need to add to make this a realistic option.


Any advice/guidance, or photos/diagrams of working compatible power supplies, appreciated!

Other helpful miscellany:

• LPL / Sauders / Omega power supplies used two different plug types to attach to the head - the older square one is a jones connector, and newer ones use a round connector. I have the older Jones. The pinout for the older connector is: Pin 1 = Red = Lamp; Pin 2 = White = Neutral (for Lamp & Fan); Pin 3 = Black = Fan; Pin 4 = Green = Ground. (source: https://www.largeformatphotography.info/forum/archive/index.php/t-12063.html and https://www.photrio.com/forum/threads/saunders-lpl-4x5-enlarger-power-supply.6235/)
• The original LPL power supplies were outsourced and produced by Hazlo (https://www.largeformatphotography.info/forum/archive/index.php/t-37563.html)

 

[koraks]

Welcome to Photrio, @brianssparetime ! Cool to see someone around who's also into electronics as well as darkroom printing

though it would not be stabilized or regulated, it’s pretty inefficient, and I have some concerns about heat dissipation.

A regular phase control dimmer is not inefficient and produces very little heat (if engineered properly). It's indeed a pretty simple/straightforward solution. Just don't inadvertently dial it at full power as you'll blow your bulb. Some form of a safeguard sounds like a good idea.

As to option 5: there's a plethora of phase control dimming circuits online that you could 'experiment' with - by this I mean initially simulate in e.g. LTspice, then maybe build one on low voltage (say 12VAC or 24VAC from a transformer) and only once you've really got that down, upgrade to the more lethal voltages.

I do know that there are 'Arduino' phase control dimmer (triac) 'modules' out there. I've never bought one, but they might be a good solution approaching #5 but with the higher-risk part already done for you. These modules generally are fed with a PWM signal or some other low-voltage input that controls the cutoff point.


But since you're actually referencing him anyway, perhaps it's good to alert @Nicholas Lindan of your question. I'm sure he has good advice to share. And I know that @albada also has an interest in electronics, although I don't know to what extent he has played with power electronics. Perhaps he might chime in as well.

 

[ic-racer]

Go with whatever you are capable of building. Or get a Chinese generic power supply for about $40 usd.

 

[brianssparetime]

Welcome to Photrio, @brianssparetime ! Cool to see someone around who's also into electronics as well as darkroom printing

Thanks, and glad someone else is here to offer advice!

A regular phase control dimmer is not inefficient and produces very little heat (if engineered properly). It's indeed a pretty simple/straightforward solution. Just don't inadvertently dial it at full power as you'll blow your bulb. Some form of a safeguard sounds like a good idea.

I was thinking a dimmer was just a big variable resistor, but I see now there are much more sophisticated versions. And agree about the danger of going over - some of the solutions I saw online using this approach put a physical stop pin on the dial.

As to option 5: there's a plethora of phase control dimming circuits online that you could 'experiment' with - by this I mean initially simulate in e.g. LTspice, then maybe build one on low voltage (say 12VAC or 24VAC from a transformer) and only once you've really got that down, upgrade to the more lethal voltages.

I do know that there are 'Arduino' phase control dimmer (triac) 'modules' out there. I've never bought one, but they might be a good solution approaching #5 but with the higher-risk part already done for you. These modules generally are fed with a PWM signal or some other low-voltage input that controls the cutoff point.

This is a great idea I had not thought of. I think I actually have a AC dimmer module for arduino (specifically this one by thewp122 on Tindie, who I think now is called RF Simplicity) in my parts bin, leftover from a previous project involving an AC motor.

As I mentioned, I'm not an expert on AC and power supplies - I was assuming that the kind of power supply to produce the AC 82v was somehow special, rather than a subset of a larger common class. Aside from "triac" and "phase control" are there other keywords/features I should be keeping an eye out for?

I was planning to test with a 200W 82V projector bulb, which are not that expensive if I blow it. I've never used LTSpice or a simulator (I did my PCB designs in EasyEDA)

But since you're actually referencing him anyway, perhaps it's good to alert @Nicholas Lindan of your question. I'm sure he has good advice to share. And I know that @albada also has an interest in electronics, although I don't know to what extent he has played with power electronics. Perhaps he might chime in as well.

Thank you!

Or get a Chinese generic power supply for about $40 usd.

I tried looking for a generic PS as described above, but I can't find any that seem like they would work to produce this kind of chopped 82v AC - most seem like fixed or variable voltage DC power supplies, which I don't think would work.

Can you provide an example, or some keywords you're using to find something at that price?

 

[ic-racer]

I think DC will power the lamp just fine.

I did rebuild a power supply for my Minolta. It used a 120V/24V transformer and when the voltage regulation circuit fried, I re-wired it with diode rectification to get the 12V for the lamp.

Minolta MOD III Lamp/Power Modification

Back in the late 70s the voltage regulated 19V power supply for my Minolta Mod III enlarger went bad. I converted it over to a 1/2 wave rectified non-regulated supply with a big diode. It has a 24V transformer, so half of that is around 13 measured volts. I used the enlarger like that all these...

www.photrio.com

 

[brianssparetime]

I think DC will power the lamp just fine.

I'm not so sure.

If it were that simple, it seems every thread would end with a suggestion to buy a variable dc boost converter and whatever ac/dc supply you need to power that.

Second, this thread (https://www.diyaudio.com/community/threads/i-need-a-82-volt-dc-power-supply.59532/) spooked me off of using DC (or at least non-pulsed DC).

I did rebuild a power supply for my Minolta. It used a 120V/24V transformer and when the voltage regulation circuit fried, I re-wired it with diode rectification to get the 12V for the lamp.

Minolta MOD III Lamp/Power Modification

Back in the late 70s the voltage regulated 19V power supply for my Minolta Mod III enlarger went bad. I converted it over to a 1/2 wave rectified non-regulated supply with a big diode. It has a 24V transformer, so half of that is around 13 measured volts. I used the enlarger like that all these...

www.photrio.com

What you did there seems like my option #2 - diode for half rectification and something to handle transient spikes. Good to know that's a viable option.


I've also read up a bit on SCR power supplies. I think the relevant difference is that SCR only chops up the positive part of the waveform and doesn't pass any of the negative, while a triac power supply can chop up both the positive and negative sides.

So if you've successfully run your enlarger off a diode that half rectifies the wave, then a SCR should work as well. Some of those seem to be quite cheap (e.g. this one on amazon).

 

[koraks]

I was thinking a dimmer was just a big variable resistor

That would be a 'rheostat' and they were used in the 'good old days' (where cozy dimmed light at the same time heated your house, yay!) They used to be pretty common, but for the drawbacks you actually pointed out, have been superseded by different solutions. Sometimes with their own drawbacks, but in principle, this will work fine for your application.

I'm not so sure.

If it were that simple, it seems every thread would end with a suggestion to buy a variable dc boost converter and whatever ac/dc supply you need to power that.

Not necessarily. And @ic-racer certainly is right - the bulb doesn't care if it's AC or DC. DC will work perfectly fine.
The reason you don't see this suggestion more often is because the advent of cheap DC supplies with high output power is fairly recent. It takes some time to trickle down into all aspects of society, and I think darkroom photography is inherently a little conservative, so maybe we take a little more time to adopt things than other people. Moreover, these units still remain much more expensive (and bigger) than e.g. a thyristor dimmer circuit. If your enlarger has, say, a 500W or 1000W incandescent bulb like some bigger Durst units, then a DC power supply still isn't all that practical. I actually discussed this very extensively with a friend a couple of months ago and I had to agree with him that it just wasn't the most obvious solution. But for modest power levels up to let's say 250W or 500W or so, it's now a viable option to use an off the shelf switch-mode power supply.

I think I actually have a AC dimmer module for arduino (specifically this one by thewp122 on Tindie

I had a brief look at it and it seems like it should be usable. I'd give it a try.

Aside from "triac" and "phase control" are there other keywords/features I should be keeping an eye out for?

I think you've already found some relevant terms. There are probably also some thyristor circuits that are usable.

Second, this thread (https://www.diyaudio.com/community/threads/i-need-a-82-volt-dc-power-supply.59532/) spooked me off of using DC (or at least non-pulsed DC).

It's from 2005. Back then, the choice of off-the-shelf SMPS units in various voltages and at low cost was much, much more limited. Today, you'd go on AliExpress etc. (if you don't mind buying straight from China), find what you're looking for and getting it shipped. In 2005, we didn't have as many options.
The main drawback is still that 82V is a bit of an oddball number in DC terms. You could look for an 80V unit and pick one with an adjustment trimmer (plenty of those around) and then dial that to the desired voltage under the actual load of the bulb. The adjustment range is usually narrow, but +/-2V will definitely be within it for an 80V unit.

 

[4season]

Option 5 (full DIY): The longer quote I pasted above as background outlines the use of a triac firing 90 degrees from the AC waveform as the basis for more sophisticated power supplies. Following up on that, I found this page (https://www.homemade-circuits.com/how-triac-phase-control-works/) (see Figure 4) that shows a power supply circuit diagram using a zener to stabilize it against input fluctuations, and a triac to produce the output wave. Interestingly, the article comments that the “set up helps to sustain virtually a constant output regardless of any significant variations in mains AC input voltage. It is regularly found in photographic and other applications where a highly stable and fixed level of light becomes essential.”

Sounds like @Nicholas Lindan figured it out back in 2008. When I owned a 4550XL, I thought it was odd that Saunders sourced the enlarger from LPL in Japan, but supplied their own USA-made power supplies. I never bothered to examine the power supply circuitry because I had the impression that it was something really simple, boring and cheap. I didn't get the sense that it contained any transformers, and certainly not computer logic. But a half-wave power supply with zener diode and a triac or two certainly might have fit inside that little box.

While DC will light an incandescent bulb as well as AC, it can be a PITA to work with at higher voltages.

 

[brianssparetime]

That would be a 'rheostat' and they were used in the 'good old days' (where cozy dimmed light at the same time heated your house, yay!)

Yes, I'm old enough and haven't lived anywhere with modern lighting, so I just assumed they still work that way.

Not necessarily. And @ic-racer certainly is right - the bulb doesn't care if it's AC or DC. DC will work perfectly fine.
The reason you don't see this suggestion more often is because the advent of cheap DC supplies with high output power is fairly recent.

Interesting. I hadn't considered the age of the comments I was reading.

I think I'm going to go with parts I already have on-hand: A C13/14 socket, a power switch, a triac ac dimmer module for arduino, a relay module (to let the timer turn off the output 82v power while leaving the fan on) wired to an old extension cord male end, and an arduino ATTiny85 to experiment with what PWM I need to get 82v on my 200w test projector bulb (and then eventually have that pwm hardcoded), and some fuses.

Sounds like @Nicholas Lindan figured it out back in 2008. When I owned a 4550XL, I thought it was odd that Saunders sourced the enlarger from LPL in Japan, but supplied their own USA-made power supplies. I never bothered to examine the power supply circuitry because I had the impression that it was something really simple, boring and cheap. I didn't get the sense that it contained any transformers, and certainly not computer logic. But a half-wave power supply with zener diode and a triac or two certainly might have fit inside that little box.

While DC will light an incandescent bulb as well as AC, it can be a PITA to work with at higher voltages.

I don't think it included a transformer, unless it was a very small pulse one.

But the box I've seen in pictures does look quite large for just having a diode OHP style power supply, and it's described as stablized and regulated, so I'm guessing it has something like a small board containing the diagram in figure 4 here (https://www.homemade-circuits.com/how-triac-phase-control-works/) with a zener to stablize and a rectifier/triac to chop up the ac waveform.

 

[albada]

But since you're actually referencing him anyway, perhaps it's good to alert @Nicholas Lindan of your question. I'm sure he has good advice to share. And I know that @albada also has an interest in electronics, although I don't know to what extent he has played with power electronics. Perhaps he might chime in as well.

I know how SMPSs work, and I've replaced blown capacitors in them. But nothing more. When I need a power supply, I buy one. I've never needed an unusual voltage though.

Mark

 

[koraks]

an arduino ATTiny85

Which will work for generating some PWM signals, but I've always found the ATtiny85 with its lack of peripherals a little awkward to work with. It has no hardware UART for instance, which makes debugging unnecessarily complicated IMO. There's a host of microcontrollers and experimenter boards out there that are packed with features - some in very small form factors.

Anyway, sounds like you're in for a fun and manageable project!

 

[4season]

But the box I've seen in pictures does look quite large for just having a diode OHP style power supply, and it's described as stablized and regulated, so I'm guessing it has something like a small board containing the diagram in figure 4 here (https://www.homemade-circuits.com/how-triac-phase-control-works/) with a zener to stablize and a rectifier/triac to chop up the ac waveform.

That looks promising!

 

[brianssparetime]

Which will work for generating some PWM signals, but I've always found the ATtiny85 with its lack of peripherals a little awkward to work with. It has no hardware UART for instance, which makes debugging unnecessarily complicated IMO. There's a host of microcontrollers and experimenter boards out there that are packed with features - some in very small form factors.

Anyway, sounds like you're in for a fun and manageable project!

The Attiny85 is indeed awkward to debug. I usually prototype with an arduino nano or something similar before recompiling for the tiny. And I never solder the tiny directly to the board - it's always socketed for easy reprogramming.

And my choice of those two is pretty arbitrary - I just happened to buy a bunch when I first got into this hobby like 10 years ago.

Curious what you'd recommend as a good more modern DIP 8/10/12 package. Not really interesting in anything that requires a vastly different toolchain or serious code rewrites.

 

[koraks]

I often fall back on 328P's, or rather 328PB's, especially if it doesn't have to go all fast and isn't web connected. Otherwise an ESP module, so 8266 or 32.

I don't do anything with DIP anymore. I know there are convenient adapter boards from various SMT packages to DIP, which opens up pretty much the entire landscape of possibilities. And if you like the Arduino toolchain (I personally certainly do), there are so many cores to choose from. Whether you like AVR and/or the many ATtiny's, Espressif, STM - lots to choose from. I mostly let myself be guided by requirements and whatever I can get my hands on

 

[brianssparetime]

I often fall back on 328P's, or rather 328PB's, especially if it doesn't have to go all fast and isn't web connected. Otherwise an ESP module, so 8266 or 32.

I don't do anything with DIP anymore. I know there are convenient adapter boards from various SMT packages to DIP, which opens up pretty much the entire landscape of possibilities. And if you like the Arduino toolchain (I personally certainly do), there are so many cores to choose from. Whether you like AVR and/or the many ATtiny's, Espressif, STM - lots to choose from. I mostly let myself be guided by requirements and whatever I can get my hands on

Interesting.

I've never done any SMT stuff. When I got into the hobby 10-12 years ago, through-hole was cheap, easy, and still very common. I'm into electronics less for becoming great than for its instrumental use towards my other hobbies, so I just standardized on a few 5v setups (nano and attiny85) to keep the learning curve manageable and compatibility broad.

Things have obviously progressed a lot since then, and I'm considering just using something like the pi pico or teensy for everything forward. But getting all the stuff to do surface mount and getting good enough to do it quickly is still a big hurdle, not to mention my entire components inventory is through-hole based.

 

[brianssparetime]

Also...

I'm starting to test and assemble my setup, and planning out the wiring for installation into a box.

I just recalled that the cord to the enlarger is 4 connectors - 120v ac hot for the fan, 82v ac for the bulb, neutral, and ground. For the neutral wire (to the enlarger), I could connect it to a) the AC in neutral, or b) the triac module's AC neutral out, or c) both.

Triac module:
()

Is one of these options more advisable? My intuition is option c, but I'm not quite sure why. Maybe this is in the land of eeoverflow?

 

[4season]

Also...

I'm starting to test and assemble my setup, and planning out the wiring for installation into a box.

I just recalled that the cord to the enlarger is 4 connectors - 120v ac hot for the fan, 82v ac for the bulb, neutral, and ground. For the neutral wire (to the enlarger), I could connect it to a) the AC in neutral, or b) the triac module's AC neutral out, or c) both.

Triac module:
()

Is one of these options more advisable? My intuition is option c, but I'm not quite sure why. Maybe this is in the land of eeoverflow?

I'd design it such that it would work safely and correctly even if the power outlet was mis-wired, with the Hot and Neutral wires reversed.

 

[brianssparetime]

I'd design it such that it would work safely and correctly even if the power outlet was mis-wired, with the Hot and Neutral wires reversed.

That obviously would be good, but it's not very instructive on how.

I am using a cord that forces the correct orientation to my correctly wired socket, but, unless there is an easy and simple way to handle this, I think I'll get it fully working first before testing to see if flipped polarity is an issue.