Printable View
Yeah, it will go more, but you'll hit a current limit before too long. At almost no restriction I was already almost getting 2.5amps draw at only 12v. I'm curious to get that variable resistor in place and see what happens. At higher restriction, I'm sure you could push 13v if heat is kept in order.Quote:
Originally Posted by xcooling
The toshiba is good far beyond 12v, but current and heat may be the constraint. I'ml do some more on this over the weekend.:)
I'm going to work on it later today. Got everything soldered onto the board yesterday. Actually my first time soldering smd components this small. Not to mention figuring out how to solder the chip. Didn't have any solder paste so I had to do it all with a soldering iron. Actually not too hard.Quote:
Originally Posted by xcooling
I made a video with a 50K variable resistor. Thats really nice to control a Laing with a 6x6mm 50K var. resistor. :D
http://www.youtube.com/watch?v=qnNWrA4o3Fc
Awesome, thanks! That 50K sounds perfect and seems to dial it way down if needed. Voltage control works too, but that resistor would be the better way to go for most people. Then you get a "DDC Strong Vario" of sorts..:)
Alright, I picked up a 50K pot at the shack for a couple of bucks. Now looking at the PCB, it appears there is a trace between Vref and Vsp. I'm assuming I need to remove this trace and connect up the pot something like this?
Yeah the way the controller works is through the voltage given to VSP. Since vref outputs 5v that puts it at max. You have to cut the trace since it'd bypass the pot.
You can see the info in http://www.toshiba-components.com/mo...20_080424_.pdf
VSP:
Duty cycle/motor speed control input (This pin has a pull-down resistor.)
0 ≤ VSP ≤ VAD (L): Sets the PWM duty cycle, based on the analog input.
VAD (H) ≤ VSP ≤ VREF: 100% duty cycle (127/128)
0 ≤ VDUTY ≤ VAD (L) → Duty cycle = 0%
VAD (L) ≤ VDUTY ≤ VAD (H) → Figure on the right (1/128 to 127/128)
VAD (H) ≤ VDUTY ≤ VDD → Duty cycle = 100% (127/128)
Since he is using the example application circuit with 100pF Cap and 20kΩ for clock generation
PWM Input Voltage:
VAD (L) = Typical is 1.2v
VAD (H) = Typical is 4.1v
I'm not sure exactly if this would work, but I would figure you could connect Vref to a resistor to ground. Then I would figure you could hook up VSP to a fan controller. It'd have to be an analog voltage though. So possibly control it via a motherboard header. Or you could make a circuit to convert a pwm signal to analog if your motherboard only had fan control via pwm. So then as the heat goes up the pump could speed up. Also could change the parameters of the pwm to analog to set a max voltage applied so it never hit max speed (keep noise down). Just a random thought.
Thanks!...It works!
Except I like it this way to make right turn an increase in speed. I tried the 50K pot and a 100K pot. The 100K would go below startup, but the 50K is just right. I used some thinner guage wire about a foot long to the 50K pot and it works out so the resistance is never too much for startup. The current is extremely low, so you could also use one of the mini pots, but they only had 100K at radioshack in the mini sizes.
12V shutoff:
Low end 2,460 RPM (VSP=5V)
High end 6,240 RPM (VSP=2.44V)
Works great..nice DDC Strong Vario..:up:
One more question, I found this:
Rotation speed output pin (open-drain)
This output is held low at startup and when an abnormality is detected. In sensorless mode,
pulses are generated at 3 ppr according to the back-EMF.
Note: 3 ppr = 3 pulses per electrical degree (With a four-pole motor, six pulses are generated
per revolution.)
The DDC is a 6 pole motor.
Electrical degrees is something completely new to me, but it appears that Hz x 10= RPM...is that right?
You and me both :D I'm an econ major. Has jack all to do with electrical or physics lol.Quote:
Originally Posted by Martinm210
I never bothered figuring out how rpms are actually read other than something like there are a certain amount of pulses per rotation. According to what it says I'd figure it's just 3ppr per 2 poles (6ppr for 4 poles) and so it'd end up being 9ppr for the DDC 6 pole motor.
It's something like
RPM / 60 = Rotations Per Second
Rotations per second * 9 ppr = Hz
So I'd figure RPM = (HZ / 9ppr) * 60
Note: Its just a guess :D
If there's an interest I can take a fan down into the lab and hook the tach up to a scope and the source up to a lab PSU and see what comes out for a fan at a known RPM :shrug: (cool gear is just one of the many perks of being an EE major :D)Quote:
Originally Posted by DumpALump
Thanks!Quote:
Originally Posted by DumpALump
That's sort of where I was too, but it seems too low. I'll just have to finally test the DC3.2 with RPM sensor then I can match them up exactly. When comparing to the DDC3.25 it seemed like Hzx10 was close, but the 3.25 has a slightly different impeller.
I never recorded RPM on a DDC3.2 test before, so I'll have to do that to be sure. Either way this will not follow the normal PC fan multiplier of Hzx30, so you'll see really high readings.
Maybe 9 is right though.
I noticed my Crystalfontz has a setting for PPR, normally 2 PPR is where it's set. Never knew what PPR meant before..Quote:
Originally Posted by m0r7if3r
But it must already have a factor of 15 built in per PPR.
I've just been using my multimeter to record Hz because it seems more accurate.
The DDC has 6 windings, does that mean 6 poles?
Us CE's shouldn't be tinkering with EE stuff....we're bound to pop a cap now and then...:D
From what I could find out, regular fans have a 2 ppr tach wire, which is what the motherboard reads. I tried looking up the voltage used, but couldn't find it
Using the calculations and a 1800rpm fan.
1800rpm / 60 = 30 rotations per second
30 rotations per second * 2 ppr = 60hz
60hz * 30 = 1800
Makes sense. Would figure if your meter is set to 2 ppr that it should correctly read fan speeds. It should be the same for the DDC pumps too since they can be read by the motherboard header.
1800rpm / 60 * 9 = 270hz
1800rpm / 270hz = 6.67 multiplier
Testing 6.67 multiplier:
4500rpm / 60 * 9 = 675hz
675hz * 6.67 multiplier = 4502rpm
So that what you are looking for?
I'm currently looking for a circuit that could convert the 9ppr to 2ppr so that it could be read by the motherboard header. I'd figure it'd be a fairly simple circuit. If it's easy enough, might be mod the diyinhk pcb for it as there is some unused space just for ground.
Here a picture with the poti at his final place. Only the rpm gives me headache.
http://www.abload.de/img/imgp0469i9js.jpg
Nice!
FYI,
I've completed a draft of my blog for the Toshiba controller version, the Sanyo is next. Since they are two different controllers, I chose to keep them separate blogs.
This is just the Toshiba:
http://martinsliquidlab.org/2011/04/...ddc-pump-mods/
http://martinsliquidlab.files.wordpr...pg?w=614&h=583
http://martinsliquidlab.files.wordpr...ng?w=548&h=602
http://martinsliquidlab.files.wordpr...ng?w=558&h=656
http://martinsliquidlab.files.wordpr...ng?w=548&h=602
http://www.youtube.com/watch?v=sRoBN...layer_embedded
Cheers!
Martin
:cool: very nice. It does sound quite a bit louder but in my opinion a good trade off for the extra performance.
Use motherboard PWM to control the Toshiba PCB modded DDC pump speed seems very simple:D
Only one resistor and capactior can do the trick!
http://hostthenpost.org/uploads/617f...e1108b02e4.jpg
anybody have time can test and let us know if it work;)
Reference: http://ww1.microchip.com/downloads/e...tes/00538c.pdf
A simple RC Filter :)
Do you have an idea how to get the rpm read by the motherboard?
Interesting, we would just need to target the 2.5 to 5.0V range, so you'd probably have to tweak the values to get it just right. One of the electrical guys on the board probably can figure it out..:shrug:Quote:
Originally Posted by wizard1238
Mine works on the crystalfontz, but RPM readout can be hit and miss on motherboards unless you also provide the ground. Try running both the signal and ground wire to a motherboard header and see if that works. That's what I have to do. It will read high if you get it to work, but it should work if you have the Tach wire soldered up.Quote:
Originally Posted by mochti01
I also can't get Hz to read via multimeter until I connect both signal and ground to a fan connector as well. Probably something about a need to drain the pulses being sent or something like that..for whatever reason it won't read unless both are connected...same for fans..:confused:
Simple electronics - you have to have a complete circuit or you get nothing ;) :D
PWM from all motherboard is 5V ~25kHz,Quote:
Originally Posted by Martinm210
50%-100% PWM = 2.5V-5V after the RC filter:p:
R=4k and C= 0.01uF is for ~20kHz (and it should work for ~25kHz) according to AN538:clap:
the next project is PWM control the sanyo board, but it's already very silent at 12V, is it need to speed down:confused:
This has been a great thread to keep tabs on. :)
I'll pour a little water on it next time, that always seems to complete my circuits..:DQuote:
Originally Posted by Sparky
Nice! Yes, PWM would be a nice addition:up:Quote:
Originally Posted by wizard1238
FYI, the Sanyo motor is nice and quiet, it sounds much more like a DDC3.1.Quote:
Originally Posted by bmaverick
Doing a quick and dirty test, not even apples to oranges. The DDC1 had an old PTS acetal top, the Sanyo had and acrylic top.
Ambient = 35.5 dbA
Ambient + DDC1 = 41.7dbA (+6.2dbA)
Ambient + DDC1 with DIYINHK Sanyo PCB = 38.1 (+2.6dbA)
I'll have to do a more apples to apples (Same Top) video recording type of test in an actual loop, but the Sanyo controller is nice and quiet.:up:
There are silent/passive builds, and there are moders which are very anal about loudness, where even silent fans @500rpm are marginally tolerable. And i recall some complaining that even DDC3.1 is too loud for them. So pwm imho won't hurt, especially as it gives easy way to regulate pump by software without purchase of expensive smart controllers.