Using IQ Cyo driver with other LEDs

I think this is a fairly simple question (famous last words...).

I am building a new dynamo light (to run from SP hub dynamo, to replace my long-serving simple but extremely effective 4x XR-E with Ledil SS optics) because none of the lights currently on the market meet what I want (still). Initial requirements are quite simple, to power 2 Luxeon Rebels in series, at 1A, preferably with a standlight (but even that is not that important). The reason for the choice of the 'old' LEDs is that I have a Philips Saferide 80 which will form the basis of the unit (case, optics and LEDs).

Rather than build a driver from scratch, it occurred to me that it might be very simple just to use a driver from an IQ Cyo, which, from research here and on the web, appears to drive a single Cree (XR-G?) at about 1A through the use of a simple courrent doubler. The standlight on the Cyo is pretty good, and appears to be just a 5.5V 1F super cap in parallel with the LED (but hard to confirm this). As far as I can tell, based on limited information, the only mod that would be needed would be to address the problem of a 5.5V cap in parallel with 2 LEDs across which the potential difference will be 6.8-7.0V. Therefore, 2 questions:

1. Is there any obvious problem with the approach described above (does anyone have any knowledge of the Cyo driver circuit)?
2. What is the easiest way to solve the problem of the super cap - simply replace the existing 5.5 1F with two 5.5V 2.2F super caps in series?

I might just go and try it anyway and see what happens (but replacing the Rebels in the Saferide 80 is a bit of pain, having already done it once...).

Thank you.

I forgot that the Saferide uses two LEDs... that does make it a nice host for a dynamo light.

Why do you want to use the Cyo circuit? Just to produce an equivalent to the 4 XR-E light? Not a bad idea. It does produce lower output at lower speeds, but should be similar to your current light.

I don't know what's in the Cyo, although I'm certainly curious. I assume it is done with something similar to a switching power supply circuit, and have sketched out a few ideas myself. Haven't found time to play with it beyond some simulations, though.

The supercap will depend on the circuit used to drive the LED. I'm not necessarily assuming that the supercap just sits at the input to the circuit, although it could. If you do just use two Cyo circuits wired in series, then I would assume that you would just have a supercap assigned to each circuit.

My question is: where will you get two Cyo circuits? If you just scavenge two Cyo's, why not just use two Cyo's wired in series?

The cyo stand light seems to get turned back on when you rotate the wheel a little, it seems to be more than a simple super cap.

Steve K said:

Why do you want to use the Cyo circuit? Just to produce an equivalent to the 4 XR-E light? Not a bad idea. It does produce lower output at lower speeds, but should be similar to your current light.

Click to expand...

The 1A current driver. My current light uses a manually switched version of Martin's voltage doubler circuit (http://www.pilom.com/BicycleElectronics/DynamoCircuits.htm), so is limited to ~0.5A. I think there are other dynamo lights that will use a similar type of driver, but I have an old cyo sitting in the garage that I could cannibalize.

Steve K said:

The supercap will depend on the circuit used to drive the LED. I'm not necessarily assuming that the supercap just sits at the input to the circuit, although it could. If you do just use two Cyo circuits wired in series, then I would assume that you would just have a supercap assigned to each circuit.

Click to expand...

I am assuming (and it is an assumption with no justification), that the cyo driver could produce 1A at a higher voltage than the 3.5V it does at the moment, with an appropriate load on the end i.e. 7V if there are 2 LEDs in series. I haven't been able to find a description which indicates that there is a voltage regulator in the circuit so I might have to take it apart and see what is in there.

Martin

drgannet said:

The 1A current driver. My current light uses a manually switched version of Martin's voltage doubler circuit (http://www.pilom.com/BicycleElectronics/DynamoCircuits.htm), so is limited to ~0.5A. I think there are other dynamo lights that will use a similar type of driver, but I have an old cyo sitting in the garage that I could cannibalize.



I am assuming (and it is an assumption with no justification), that the cyo driver could produce 1A at a higher voltage than the 3.5V it does at the moment, with an appropriate load on the end i.e. 7V if there are 2 LEDs in series. I haven't been able to find a description which indicates that there is a voltage regulator in the circuit so I might have to take it apart and see what is in there.

Martin

Click to expand...

a quick google search didn't pull up any schematics for the Cyo, although it did pull up this thread and a variety of other bits that I recognize (including my Reddy Kilowatt avatar).
My guess would be that the parts in the Cyo could handle the output voltage without hurting it, but the input voltage would have to double too, and that might interfere with its operation (just my guess).

I'd be interested in seeing how much could be reverse engineered from the Cyo. It could take a bit of work, especially since the small parts usually don't have any markings that suggest what they are. Got any photos of the board?

Steve K said:

a quick google search didn't pull up any schematics for the Cyo, although it did pull up this thread and a variety of other bits that I recognize (including my Reddy Kilowatt avatar).
My guess would be that the parts in the Cyo could handle the output voltage without hurting it, but the input voltage would have to double too, and that might interfere with its operation (just my guess).

I'd be interested in seeing how much could be reverse engineered from the Cyo. It could take a bit of work, especially since the small parts usually don't have any markings that suggest what they are. Got any photos of the board?

Click to expand...

There are pictures of a tear down here http://fotos.mtb-news.de/s/12234. As you wrote, it is quite hard to work out the circuit, so I just had a go at running the Saferide 80 from it. The summary is that it works very well.

Initially running without the standlight super cap produced a bit of flicker up to about 10 km/h, but then a smooth bright light with about 0.9A delivered at about 30 km/h. There was a gradual increase in brightness up to 45 km/h at which point there was the best part of 1A delivered (current was estimated from brightness measurements, compared to the light running on battereis with a known current of 0.7A). Voltage dropped across the LEDs was a bit over 5.9V at 30 km/h and just under 6V at 50 km/h (Rebel datasheet gives 3V at 1A).

Adding 2 super caps in series in place of the single 5.5V 1F cap gave a reasonable standlight, but more importantly reduced the low speed flicker significantly (on the cyo the standlight kicks in at about 7.5 km/h, and is one of the most flicker free lights I have come across).

Unfortunately, jsut as I was making a final test before sealing everything up in the light, a flying lead to the LEDs made contact with the MCU on the driver circuit and it no longer delivers 1A! All I can get is a rectified direct drive at 0.5A. Ho hum, that will teach me to be a bit more careful.

I'm keen to have another go, now I know it works. What do you reckon is the cheapest light with a decent currect driver to try? Herrmans H-Diver? Might just go for the cheapest cyo..

drgannet said:

so I just had a go at running the Saferide 80 from it. The summary is that it works very well.

Click to expand...

See? That's what I would have bought from Philips. They diluted their resources by getting into all sorts of inferior derivatives.
And thanks for donating that board, Doc!

drgannet said:

There are pictures of a tear down here http://fotos.mtb-news.de/s/12234. As you wrote, it is quite hard to work out the circuit, so I just had a go at running the Saferide 80 from it. The summary is that it works very well.

Click to expand...

I guess I had seen those photos before. Looking at it, the 14 pin IC looks like a T.I. switching controller, perhaps, and the 8 pin SOIC looks like it is probably a mosfet. One photo suggested that the switching controller's part number is CM01068, or maybe CM0106888?? I couldn't get any matches on the TI web site.

Sorry to hear that you killed your part. If you, or someone, can figure out what that T.I. part is, it might be a simple matter of replacing it. My guess is that the buck regulator is just stuck with the mosfet turned on.

Marcturus said:

See? That's what I would have bought from Philips. They diluted their resources by getting into all sorts of inferior derivatives.
And thanks for donating that board, Doc!

Click to expand...

Yes, the StVZO regulations have a lot to answer for - it would be so easy for a large light manufacturer to make something like this!

Steve K said:

I guess I had seen those photos before. Looking at it, the 14 pin IC looks like a T.I. switching controller, perhaps, and the 8 pin SOIC looks like it is probably a mosfet. One photo suggested that the switching controller's part number is CM01068, or maybe CM0106888?? I couldn't get any matches on the TI web site.

Sorry to hear that you killed your part. If you, or someone, can figure out what that T.I. part is, it might be a simple matter of replacing it. My guess is that the buck regulator is just stuck with the mosfet turned on.

Click to expand...

The part number is CM0106B and underneath is written 09K and under that CSL2. TI make one of those but not in the same package.

However, I think it was the 8 pin IC that contacted with the LED output lead, so if it is just that that has gone it could be easier to replace rather than the whole driver. You are right Steve that it is a mosfet. The part number is very hard to read but I think it is 4816N (30V, 11A n-channel mosfet).

Martin

drgannet said:

The part number is CM0106B and underneath is written 09K and under that CSL2. TI make one of those but not in the same package.

Click to expand...

Do you have a link to this T.I. part? I'm having no success searching for it on ti.com or digikey.com.

Steve K said:

Do you have a link to this T.I. part? I'm having no success searching for it on ti.com or digikey.com.

Click to expand...

I didn't follow the link through, but it doesn't look like the part - FWIW this was what it was pointing to http://www.ti.com/product/cd40106b

yeah, that came up when I searched too. It's a nice part.. the 4000 series CMOS logic family is really handy for a lot of things and has been in production for 30 years or more. It's not a switching power supply controller, though.

I'm wondering if the CM0106 gadget might even be a microcontroller or programmable logic device. It's frustrating because that sure looks like the T.I. logo on the part, but that part number just isn't showing up in a search.

Well, I tried a repair on the cyo circuit but it appears I destroyed more than the mosfet because replacing that made no difference. Going to try the circuit from an Axa Pico 30-T next, as I can get hold of one of those pretty cheaply. Watch this space...

drgannet said:

What do you reckon is the cheapest light with a decent currect driver to try? Herrmans H-Diver?

Click to expand...

CPF member SWHS have written that he has used a Herrmans H-diver driver.
http://swhs.home.xs4all.nl/fiets/tests/verlichting/index_en.html
http://www.bherrmans.fi/HEADLEDLIGHTH-DIVERDYNAMO.html

A Herrmans H-diver light is 23,36 € at Rose
http://www.rosebikes.com/article/herrmans-h-diver-front-headlamp/aid:553912

OK, here's the next update, on the Axa Pico 30-T, which arrived yesterday. Unfortunately it is not suitable for what I am looking for. There is no kind of current regulation on the driver, and it seems to consist solely of a rectifer with supercap across the output as a standlight - certianly nothing to give me more than about 0.5A. I thought this might be the case for these kind of very low cost but low Lux lights, but it's hard to find any information. Anyway, I am reluctant to try any more (even though Wouter does report success with the Herrmans H-diver driver I can't find any details of the output compared to other lights). Therefore, I've bitten the bullet and ordered another cyo (with some other bits to make up the order to get free postage, the cost isn't horrendous at 30-somthing Euros). Bit of a shame I only want the driver, but I couldn't persuade the nice people at B&M to sell/send me just a circuit board. If I can get it working well with the Philips light I'll post some photos...