HMv4.0 to HMv4.1?


 

RalphTrimble

TVWBB Diamond Member
One of the actual electronic differences between the HMv4.0 and HMv4.1 boards is the way the blower is regulated. HMv4.0 pulses the ground to regulate speed, HM4.1 pulses the 12v side instead, which simplifies wiring cause the servo and blower can now share a ground.

This, however, makes the wiring for the add-in CAT5 jack on HMv4.0 different than the HMv4.1 and also limits the number of probes that can be connected to two instead of three.

I am wondering (Bryan) if you think it would be possible to modify the blower driver circuit on a HMv4.0 board and change the MOSFET to the type used on HMv4.1 so the CAT5 wiring on a HMv4.0 board (and blower operation) can match the standard HMv4.1 pinout?
 
It is possible, but it requires a bunch of rework because there's so much different about it. Also there's no place to fit the new components needed so it would all have to be soldered together deadbug style and then attached to the PCB. Here's a rundown of what would need to be done.

-- Remove the old MOSFET completely
-- Short the old MOSFET's drain to source
-- Attach old MOSFET gate to BC337 transistor base
-- Wire BC337 emitter to ground
-- Remove 25V/22uF blower capacitor
-- Cut the 12V PCB trace going to the blower output
-- Wire 12V to one end of a 10k resistor, and the drain (or is it called the source) of the new MOSFET, and the collector of the BC337
-- Wire remaining new MOSFET pin to old blower output, and blower capacitor +
-- Wire blower capacitor - to ground

So it's not exactly trivial but can be done and I *think* that's it.
 
Damn, I was hoping it would be easier... I will have to digest what you have posted a bit before I decide if I will do the surgery... Thanks for the detailed answer Bryan.
 
PS Where are you at these days with the V4.1 board debugging? Have you nailed down any of the noise issues? How likely is it there will be a v4.2 board in the near future?
Instead of doing the above mods to the v4.0 board I could just build up a new HM board, I have most of the parts here laying around, but I've been trying to hold out for the next release. I am putting together a unit to send to my uncle, I was going to send my v4.0 board but wanted the CAT5 wiring to match the v4.1 standard so when I send him damper samples I don't have to do special v4.0 wiring on them (and I think he would like to use all three probes over the CAT5 cable).
 
OK, digging through schematics to compare and formulate a possible go plan for the v4.0-v4.1 transition and it seems more doable than I originally thought. Let me run this plan of attack by you to see what you think....

First, remove the RCA jack from the HMv4.0 board, only CAT5 jack will be used for the blower like in v4.1.

Next, lift the transistor side of R3 from the board. This basically leaves the old blower driver circuit isolated so I should be able to leave it alone and onboard.

Next, construct the v4.1 driver circuit consisting of (from the V4.1 schematic):
Q1 (BC337), R14 (10K), Q3 (FQU11P06TU), D2 (1N4001), and C12 (22uf/25v)

Finally, tap the above (isolated) circuit to gnd and 12V and connect the input to the lifted end of R3 and output to the CAT5 jack.

I assume after removing the RCA jack and lifting R3 I will not have to cut any traces on the v4.0 board or remove any of the v4.0 driver components from the board, as long as the v4.1 circuit is complete (as above) and isolated...

Does that sound right to you? If the above conversion is doable I will probably mod my v4.0 board and keep it and send my V4.1 board to my uncle instead, I'd rather keep the frankenstien board for myself in case something goes wrong....
 
Last edited:
Yup you could do it that way too, assuming when you say "lift the transistor side of R3" you mean the MOSFET side. It isn't a terribly complicated circuit but you will have to build it all up over the existing parts.

I've been working on a (another) v4.2 board that's completely redesigned from the connector footprint side. The idea is to keep the blower output, power input, and 5v section together and put the thermistors jacks, atmega and thermocouple connector on the "far end" so there's no return current passing through that section. I haven't been able to fit it all together, and have resorted to making paper cutouts of the Pi, LCD, and heatermeter to try and even rearrange the basic layout of those. I haven't been very successful but I haven't had much time for it as every weekend since February I've had house guests.

You guys will be the first to know when the next test board order goes out. After that you're looking at probably a month if it works, longer if it doesn't. That's still a ways off too (a nebulous distance depending on the design).
 
OK, thanks for the reply. I am almost done with building the v4.1 blower circuit already, should have it installed and tested by this evening. That will hold me over for now since the boards I have will all be up to speed (v4.1 compatible CAT5 connection), I still have a spare v4.0 board I can use if I need another. Whenever you get the next generation board ready I will jump in on that...
Thanks again for your reply and all your hard work developing the Heater Meter...
 
So that worked out great, I followed the steps I outlined above and my HMv4.0 is now running the HMv4.1 blower driver circuit! I took advantage of the space created when I removed the RCA jack and installed my new circuit there, it fit in nice and neat. I took some pics along the way in case anyone else wants to do this mod to their HMv4.0.

First off, here is the HMv4.1 blower driver schematic, the circuit that needs to be built and also the pinout for the CAT5 jack:
HMv4.1_BlowerCircuit_CAT5Pinout.jpg


To construct the circuit I started out with three groups of components.

I bent back the center leg (Base) of the BC337 transistor over the flat portion, that is the input that comes from R3 on the HMv4.0 board.
I bent up the top leg (Emitter) of the BC337 transistor, this is going to connect to GND

I bent the center leg of the MOSFET (Drain) over the top of the body,
I bent the bottom leg of the MOSFET (Source) down
I installed the 10K resistor between the outer two legs of the MOSFET (Source and Gate)

I connected the 22mf/25V cap across the 1N4001 diode (mind polarity)

Here is a picture of those three basic assembly's:
4-4.1Parts.jpg


I then inverted the BC337 assembly, trimmed the non-bent leg (Collector) a bit and soldered it to the top lead (Gate) of the MOSFET

I then connected the Cap/Diode pair between the Drain (Cap + side) of the MOSFET and the Emitter of the BC337, leave the leg long cause it will solder to the HMv4.0 board gnd.

I then added the input wire to the BC337 base which will connect to the lifted side of R3

Then added the ground wire (Black) for the blower, and the +12v_BLW wire (Red) for the blower connected to the MOSFET Drain and did a little shrink wrapping of the circuit.
4-4.1Circuit.jpg


Finally, I trimmed up the excess legs leaving enough to connect to the HM board. The +12v leg can go down the original hole for the blower +12V, which is also marked +12v on the back of the board. The gnd leg can connect to the bottom leg of the existing (HMv4.0) MOSFET, 'cause it is gnd. The circuit should fit neatly into the space the RCA jack used to occupy. Then simply connect the input wire to the lifted leg of R3, and wire your new blower leads to the CAT5 jack according to the diagram.

Here is a picture of the circuit installed on the board:
4-4.1Installed.jpg


I fit it together, closed the case and tested her out. Fits nicely in the case and works as expected, SUCCESS!
 
Last edited:
Ralph,

Nice writeup.

I'm currently at the same place, building a 4.1 and have a working 4.0 setup. I was just thinking about the differences. I didn't realize that the 4.0 pulsed ground to regulate the fan speed and probably would have wound up scratching my head for hours after not realizing why my ground join mod inside of the servo/blower I had planned would have probably left the fan on at full, all of the time!

Actually, now that I think about it, since I'm not using the cat5 for probes, I can probably leave everything wired the same on my 4.0 and servo/blower unit and just add a second ground to the cat5 jack on the 4.1 board to be a dedicated ground for the fan.
 
Sorry to bump an older thread like this, but I would like to add a success story. Ralph helped me source the extra parts, a servo-barrel-case and iron out the last problems I was having. I did a straight mod from HMv4.0 to v4.1, just finished it and updated the software to v12.

Additionally to the steps outlined above, I connected:
  • the ground of the mod to the MOSFET on the board, using the connector closest to the edge of the board.
  • the +12VDC of the mod to the removed RCA connector on the board, just find the marking +12VDC and your good.
  • the +5VDC for the servo on the ethernet connector to "any marked +5VDC" on the board (I used the +5VDC on the RPI interface).
  • the SERVO signal from Pin 14 on the ATMega. Just have a look at Servo-Mode on the wiki, the pictures show the signal pin.
  • plus three probe connections from the front of the board and the accompanying ground connection next to it! Very important to use a separate ground connection for the probes.

This in total makes 8 connections and thus I'm using the full ethernet capacity. I slotted the ethernet connector in the existing case I have and it fits tightly but surely. \o/ I plannend on making pictures but I couldn't resist firing up the grill immediately for a test run. I'll be outside for a while!

Thanks again Ralph!
 
Ralph, Installed your blower mod and added an RJ45 jack for blower, servo and 3 probes to my 4.0. Using your idea of running cat5 to deck I also installed the 3 probe jacks on my blower. It is working great except that I am getting the noise icons and the pit probe is jumping around a couple degrees, the food probes are only jumping 1 degree. I am running about 30 ft of cat 5 with two wall jacks, one in house and other on deck. Not sure if noise is due to just length of wire or I need to maybe work on connections in wall jacks, solder instead of current press in connections. Also could be just because I'm testing with blower and probes at low 30's temperature. Are you getting noise when you run your 4.0 over the cat5? Thanks again for these mods, they really put new life in a 4.0! Dan

Heatermeter_zpsac7b1dd2.jpg
[/URL][/IMG]
 
Well, soldered the wires in the wall jacks (sort of), I stripped the ends and tinned them then pushed them back in, no help. Also used heat gun and ran pit up to 230 also no help. If I hear that the 4.0 probes can run over cat5 without noise I guess next step would be to try solid wire cat5 and do a real solder job to wall jacks. I know the 4.0 doesn't have the noise suppression circuit of the 4.2 so maybe this is as good as it gets. Will try a real live cook soon and see how the 2 degree oscillations affect the cook.
 
Dan, Glad you got your HMv4.0 up to speed with these mods....

On the probe noise, I can not answer your question directly because I haven't run my HMv4.0 with the latest firmware that shows the noise indicator, but my HMv4.0 probes did read very stable (better than my HMv4.1 board).

Couple things to look into on the probe noise. First GROUND.... The probes need their own ground wire, if you share the ground from the servo/blower you will get noise on the probes. It also matters which wires in the CAT5 cable you use for the probes, use twisted pairs for the servo/blower power and also twisted pairs for the probes. I tend to use a twisted pair for the Pit Probe and probe ground, then use a another twisted pair for the next two probes. You may be right about your punch down wires depending on the CAT5 wire you have, you definitely want SOLID wire not stranded for punch down connectors. I had initially trimmed away the plastic and soldered the wires to the CAT5 jack, lately I punch down all the wires but trim and solder the ground wire(s) to the connector. Finally, you should pay attention to the routing of the wires you jumper from the CAT5 jack to the probe jacks on the HM. I route the wires directly off of the board from the CAT5 jack and then run them (still in twisted pairs) along the outside edge of the board to the probe side, then make a sharp right angle up to connect to the probes. If the wires run to close to some of the switching components they can pick up noise. Speaking of switching, I THINK with the HMv4.1 blower circuit you can run the blower in VOLTAGE mode, which is less noisy than pulse mode. (I say I think because Bryan added a feedback circuit to the blower on the HMv4.2 and the Voltage mode came out around then and I am not 100% sure it works with the HMv4.1 blower driver).

Finally, and probably most important. You can easily add the RC filter(s) Bryan added to each probe on the HMv4.2 to your probes that come in on the CAT5 jack, these filters do a great job of removing the noise from the probes. Just put a 100K resistor in series with each probe lead and then connect a 0.1uf cap from each probe lead to ground. Look at the HMv4.2 schematic where the probes enter the ATMega, your HM has only the 10K pullup resistors shown there, the 100K resistor and 0.1uf caps are missing. I always connect my CAT5 probe leads to the SWITCHED leg of your onboard probe jacks (similar to how you would wire the ambient temp thermistor). This way the CAT5 probes work by default but the onboard probe jacks will work instead if you plug a probe directly into the HM. This is a better wiring scheme 'cause it eliminates the possibility of having two probes (or probe jacks and extra wires) connected in parallel
 
Last edited:
Thanks for all the ideas! I'll run through them all and report back. I did try voltage mode when I first made the mod and it doesn't work. The fan will come on when it's supposed to but always at 100%, even when a lower max is specified. No control of fan speed even in manual mode. Works fine in pulse mode. Thanks again.
 
Yup, voltage mode only works if you have a v4.2 board. It's a special circuit on the board that makes it work so voltage mode always has random output (usually full on).
 
OK, thanks for confirming that... I will have to look into constructing an add-on feedback circuit too... shouldn't be too tough...
 
I am on my phone so I can't draw a picture, but you need the feedback circuit and also the big inductor between the diode and the capacitor on the mosfet low side.
 
I am on my phone so I can't draw a picture, but you need the feedback circuit and also the big inductor between the diode and the capacitor on the mosfet low side.

I appreciate that Bryan, I realize I need to add that inductor and the rest of the feedback circuit. I tried last night to add the inductor and everything on to the HMv4.1 blower driver mod I had made (which has a different pre-driver transistor, a different cap and diode, and the cap ends up on the wrong side of the inductor). The feedback circuit did not work when wired that way so I am going to start fresh and build the HMv4.2 blower driver circuit complete and install that and see how it goes. I will post details here if/when I get it working....
 
The recent revival of this thread brought to mind that the blower circuit has been updated again on the HMv4.2. This time Bryan added a feedback circuit to the ATMega which allows the option to run the blower in the new (less noisy) Voltage Mode (as apposed to the Pulse mode used on previous HM releases which can cause more noise on the probes).

I first tried to add the extra components onto the mod detailed previously in this thread, which was unsuccessful because there are quite a few component changes in those parts as well as the addition of the feedback circuit. I can't see an easy way to update the earlier mod so I tossed it aside and started building the HMv4.2 blower circuit fresh.

First of all, here is the HMv4.2 blower driver schematic with some notes added to make wiring it up easier:
HMv4.2_BlowerDriver.jpg


Here are the components set together in basic blocks:
HMv4.2_BlowerDriverParts.jpg


Here is the circuit completely assembled:
HMv4.2BlowerCircuit.jpg


Here is the circuit installed in my HMv4.0 in the place where the RCA jack used to be, with the feedback circuit tucked behind the powe jack:
HMv4.0-4.2.jpg

It's snug but fits in there nicely....

Here is the board placed back into my case:
HMv4.0-4.2Mod.jpg


The case closed up fine and best of all IT WORKS! Now I can use Voltage mode on the blower on my HMv4.0...
 
Last edited:
Great! I ordered the caps and 100k resistors today and will try that first and if I still get noise will replace with updated blower circuit.
 

 

Back
Top