Handset CAD anyone? :)

Genesis · 5th January 2006, 15:10

Yes, I've considered a current mirror, but it has undesirable qualities, the most important of which is the continuous current dissipation required on the pass transistor (equal to the solenoid dissipation, or 1.5W continuous.) This is beyond many if not most TO92 package transistor ratings and if potted may even be beyond a TO220 (since there's no external heat sink possible.)

The switching transistor I'm using for output is a NTE11; this is a particularly nice transistor to use for this sort of application, as it has a high Hfe (~200) - thus, a darlington configuration is not required to drive it to saturation even with a very low (1 ma) base current. This avoids heating the transistor during normal operation as its dissipation is limited to the C-E voltage drop and is very friendly to the uP's drive circuits. The NTE11, for example, has a dissipation limit of 750mw - in regulating mode it is unsuitable but when saturated its fine as its maximum current dissipation under load is given by 0.7 * (1.5W / 7.5V - maximum battery supply fully charge) = 140mw - well under its rating.

This also, however, makes it very succeptable to damage from shorts, thus necessitating protective measures of some form....

LS_DK · 6th January 2006, 13:27

See, this is why I like these sort of forums.. You always learn something new.
I was very surprised to read how much power you expected to use in your Solenoid.
I had based my design on what the SnapTite folks require for their Solenoid. I wanted to drive the Inspiration Solenoid which works at a lower IP than the Snap-tite, and it was custom made for the inspiration. So I assumed the 100mA on 6 V required for the Snaptite would more than enough for my Inspiration Solenoid.
Yesterday I pulled my head apart and measured the characteristics of the solenoid.. It uses a whopping 350 mA on 6 Volt (5.87V) to stay open. (With 6 bar IP).. That is roughly 2 Watts.. That calls for a smilie,, Here you go : ... Absolutely useless.. I have spent hour upon hour optimizing for uAmps to get good battery life..
With this sort of power consumption, I agree current mirrors are not the way forward..
If I was ever going to do an interface for this sort of solenoid I would go for an integrated chip from e.g. Maxim, that does the voltage control for you and have internal thermal shut down and current limiting as well.

Thanks a lot,
Kasse

Genesis · 6th January 2006, 14:49

I considered a chip designed for servo control but it's a 14 pin DIP and has 4 channels - three more than I need

I'll have to poke around Maxim's stuff....

I have designed around a 1.5W coil, as while the 600mw coil from SnapTite is available, the smallest orifice isn't any more. That means that your maximum rated pressure on that solenoid is only 110psi - that's too close to the margins for me and an overbalanced 1st stage - such as the DS4s I prefer - could have trouble deep.

My intent here is for the controller to work with a solenoid of the user's choosing with up to a roughly 250ma current draw - or a 1.5W coil - with enough safety margin to be able to run the power a bit higher if you need to for some reason.

AD_ward9 · 6th January 2006, 23:04

Quote: (Originally Posted by Genesis)

The switching transistor I'm using for output is a NTE11; this is a particularly nice transistor to use for this sort of application, as it has a high Hfe (~200) - thus, a darlington configuration is not required to drive it to saturation even with a very low (1 ma) base current. This avoids heating the transistor during normal operation as its dissipation is limited to the C-E voltage drop and is very friendly to the uP's drive circuits.

Use a FET and a diode (to prevent the back EMF from zapping the FET), as the drop across the FET will be much less than a bipolar device, and you waste no current driving it. You cannot get rid of the heat, so using a bigger transistor does not help you: best to avoid the heat in the first place. Solid state relay FETs are suitable, some have on resistance of a few milli ohms. You may also want to have a look at the circuits we posted for this, which handle both steppers and single solenoids. The circuits are on http://www.deeplife.co.uk/or_files/F...8_Complete.zip and are for a whole rebreather. Use only Circuit set B (set A was used for the review to highlight some common mistakes - the chap who drew them got a bit enthusiastic about the O.R. concept).

If you are worried about current drain, the usual trick is put in a decent capacitor in series with a very low value resistor to give the current to move the solenoid, with a higher value resistor in parallel, to hold the coil in.

What batteries are you using? The only ones we found stand fast pressurisation and depressurisation are Lithium Ion Gel rechargeables. There are some quite nasty reports available for some other batteries in common use in rebreathers, one particular report by the US Navy comes to mind. Li Ion gel gives you tons of power. Charging circuit is in the above mentioned circuits also.

Cheers
Alex

Genesis · 6th January 2006, 23:25

Quote: (Originally Posted by AD_ward9)

Use a FET and a diode (to prevent the back EMF from zapping the FET), as the drop across the FET will be much less than a bipolar device, and you waste no current driving it.

An FET works too.

Quote:

You cannot get rid of the heat, so using a bigger transistor does not help you: best to avoid the heat in the first place. Solid state relay FETs are suitable, some have on resistance of a few milli ohms. You may also want to have a look at the circuits we posted for this, which handle both steppers and single solenoids. The circuits are on http://www.deeplife.co.uk/or_files/F...8_Complete.zip and are for a whole rebreather. Use only Circuit set B (set A was used for the review to highlight some common mistakes - the chap who drew them got a bit enthusiastic about the O.R. concept).

I'll take a look - the issue remains short circuit protection though. There's no heating problem with the NTE11 in normal operation - the issue only comes about if there's a fault.

Quote:

If you are worried about current drain, the usual trick is put in a decent capacitor in series with a very low value resistor to give the current to move the solenoid, with a higher value resistor in parallel, to hold the coil in.

No, no problem there at all actually..... the current draw is not significant, other than its effect on battery life.

Quote:

What batteries are you using? The only ones we found stand fast pressurisation and depressurisation are Lithium Ion Gel rechargeables. There are some quite nasty reports available for some other batteries in common use in rebreathers, one particular report by the US Navy comes to mind. Li Ion gel gives you tons of power. Charging circuit is in the above mentioned circuits also.

Cheers
Alex

The batteries are going in a light can along with the main processor board and are both at 1 ATA and completely isolated from the breathing loop. I'm aware of some of the problems that can arise with batteries in a breathing loop or which have potential access to it.....

LiIon gives you tons of power but also comes with some serious management issues and risks, as they are very ill-behaved if abused. I don't need that sort of power; NiMH works just fine. I'm getting plenty of runtime off a pack of 5 AA NiMH rechargeables - with the backlight off and no injection it approaches 100 hours before the low battery warning level is reached. With a "realistic" operational profile I'm seeing roughly 20 hours continuous use - vastly more than you can put on the unit during a diving day.

WaTeRwEaSeL · 7th January 2006, 02:23

Genesis,
The handset looks great. Just a quick question on how you imported your CAD files onto the board. Did you transfer them to JPEG? I have a few CAD files I would like to post and I was just wondering. (Im using AutoCAD)
Thanks

Justin

Genesis · 7th January 2006, 02:27

I "printed" them to postscript (to a file) and then used Photoshop to convert to JPEG.

I can also do a direct JPEG export from my CAD software if I want; it depends on what sort of detail you're after - the postscript conversion produces a much finer-grained image since its intended for printing at 300-600 dpi.

If your software doesn't have a direct export the printing trick works with basically any package - just set up a "printer" that's a postscript to file output (use any common printer postscript driver under windows) and you're all set.

Genesis · 25th January 2006, 05:49

I've decided to pull the posts from the lid (it makes it MUCH easier to machine and the base material is much cheaper) as I found stand-offs that will work perfectly.

Thus, the LCD will be mounted using nylon standoffs and machine screws to the lid, meaning the lid requires only the "lip" be machined and the holes drilled.

I'll post an updated exploded view (I've also got the holes located for the magnetic switches as I've done a lot of testing now with both hall effect sensors and reed switches) soon - I've been tied up with boat work the last couple of weeks.....

I'm waiting for some info from a couple of connector manufacturers before deciding whether to do the "entry hole" as a cable gland (easy and cheap) or use a connector. The connector (e.g. a fischer or one of the other competitors) has certain advantages, the most important being that a cable breach cannot propagate into the handset since the cable is "interrupted" at the connector.....

Scuba_Vixen · 25th January 2006, 14:10

Quote: (Originally Posted by Genesis)

~snip~
I'm waiting for some info from a couple of connector manufacturers before deciding whether to do the "entry hole" as a cable gland (easy and cheap) or use a connector. The connector (e.g. a fischer or one of the other competitors) has certain advantages, the most important being that a cable breach cannot propagate into the handset since the cable is "interrupted" at the connector.....

I went with connectors for mine, for exactly the cable-breach/snag reason. Any problem with a cable and the controller remains unaffected. Cables are a lot cheaper to replace than a controller.

As I posted previously, I liked the ones from Woodhead Connectivity, made by Brad Harrison. These are the same ones that Gene uses for the HS PO2 monitor.

Well imagine my surprise to find what appeared to be the same ones in McMaster Carr, though with no mfg. mentioned. (I had been having a hard time sourcing them in small quantity) So I ordered a couple, and they are indeed the right ones, complete with the Woodhead identifcation and part numbers.

I used the nano-change size for the HUD display, and the micro-change size for the cell holder. Would have liked the nanos for both, but had to work with a 20mm thread on the cell holder.

Page 721 on McMaster site to see them

I'll post a few pics when I get home from work and have time.

Darlene

Genesis · 25th January 2006, 15:07

Cool - thanks..... will look into those....

5th January 2006, 15:10	#21 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,453	Re: Handset CAD anyone? :) Yes, I've considered a current mirror, but it has undesirable qualities, the most important of which is the continuous current dissipation required on the pass transistor (equal to the solenoid dissipation, or 1.5W continuous.) This is beyond many if not most TO92 package transistor ratings and if potted may even be beyond a TO220 (since there's no external heat sink possible.) The switching transistor I'm using for output is a NTE11; this is a particularly nice transistor to use for this sort of application, as it has a high Hfe (~200) - thus, a darlington configuration is not required to drive it to saturation even with a very low (1 ma) base current. This avoids heating the transistor during normal operation as its dissipation is limited to the C-E voltage drop and is very friendly to the uP's drive circuits. The NTE11, for example, has a dissipation limit of 750mw - in regulating mode it is unsuitable but when saturated its fine as its maximum current dissipation under load is given by 0.7 * (1.5W / 7.5V - maximum battery supply fully charge) = 140mw - well under its rating. This also, however, makes it very succeptable to damage from shorts, thus necessitating protective measures of some form....
(Offline)

6th January 2006, 13:27	#22 (permalink)
LS_DK New Member Current Rebreather/s: Inspiration Classic Other Rebreather/s: Join Date: Mar 2005 Location: London, UK Posts: 37	Re: Handset CAD anyone? :) See, this is why I like these sort of forums.. You always learn something new. I was very surprised to read how much power you expected to use in your Solenoid. I had based my design on what the SnapTite folks require for their Solenoid. I wanted to drive the Inspiration Solenoid which works at a lower IP than the Snap-tite, and it was custom made for the inspiration. So I assumed the 100mA on 6 V required for the Snaptite would more than enough for my Inspiration Solenoid. Yesterday I pulled my head apart and measured the characteristics of the solenoid.. It uses a whopping 350 mA on 6 Volt (5.87V) to stay open. (With 6 bar IP).. That is roughly 2 Watts.. That calls for a smilie,, Here you go : ... Absolutely useless.. I have spent hour upon hour optimizing for uAmps to get good battery life.. With this sort of power consumption, I agree current mirrors are not the way forward.. If I was ever going to do an interface for this sort of solenoid I would go for an integrated chip from e.g. Maxim, that does the voltage control for you and have internal thermal shut down and current limiting as well. Thanks a lot, Kasse
(Offline)

6th January 2006, 14:49	#23 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,453	Re: Handset CAD anyone? :) I considered a chip designed for servo control but it's a 14 pin DIP and has 4 channels - three more than I need I'll have to poke around Maxim's stuff.... I have designed around a 1.5W coil, as while the 600mw coil from SnapTite is available, the smallest orifice isn't any more. That means that your maximum rated pressure on that solenoid is only 110psi - that's too close to the margins for me and an overbalanced 1st stage - such as the DS4s I prefer - could have trouble deep. My intent here is for the controller to work with a solenoid of the user's choosing with up to a roughly 250ma current draw - or a 1.5W coil - with enough safety margin to be able to run the power a bit higher if you need to for some reason.
(Offline)

6th January 2006, 23:04	#24 (permalink)
AD_ward9 RBW Member Current Rebreather/s: Other CCR Other Rebreather/s: Other CCR Join Date: Jun 2005 Location: Scotland Posts: 2,011	Re: Handset CAD anyone? :) Quote: (Originally Posted by Genesis) The switching transistor I'm using for output is a NTE11; this is a particularly nice transistor to use for this sort of application, as it has a high Hfe (~200) - thus, a darlington configuration is not required to drive it to saturation even with a very low (1 ma) base current. This avoids heating the transistor during normal operation as its dissipation is limited to the C-E voltage drop and is very friendly to the uP's drive circuits. Use a FET and a diode (to prevent the back EMF from zapping the FET), as the drop across the FET will be much less than a bipolar device, and you waste no current driving it. You cannot get rid of the heat, so using a bigger transistor does not help you: best to avoid the heat in the first place. Solid state relay FETs are suitable, some have on resistance of a few milli ohms. You may also want to have a look at the circuits we posted for this, which handle both steppers and single solenoids. The circuits are on http://www.deeplife.co.uk/or_files/F...8_Complete.zip and are for a whole rebreather. Use only Circuit set B (set A was used for the review to highlight some common mistakes - the chap who drew them got a bit enthusiastic about the O.R. concept). If you are worried about current drain, the usual trick is put in a decent capacitor in series with a very low value resistor to give the current to move the solenoid, with a higher value resistor in parallel, to hold the coil in. What batteries are you using? The only ones we found stand fast pressurisation and depressurisation are Lithium Ion Gel rechargeables. There are some quite nasty reports available for some other batteries in common use in rebreathers, one particular report by the US Navy comes to mind. Li Ion gel gives you tons of power. Charging circuit is in the above mentioned circuits also. Cheers Alex
(Offline)

6th January 2006, 23:25	#25 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,453	Re: Handset CAD anyone? :) Quote: (Originally Posted by AD_ward9) Use a FET and a diode (to prevent the back EMF from zapping the FET), as the drop across the FET will be much less than a bipolar device, and you waste no current driving it. An FET works too. Quote: You cannot get rid of the heat, so using a bigger transistor does not help you: best to avoid the heat in the first place. Solid state relay FETs are suitable, some have on resistance of a few milli ohms. You may also want to have a look at the circuits we posted for this, which handle both steppers and single solenoids. The circuits are on http://www.deeplife.co.uk/or_files/F...8_Complete.zip and are for a whole rebreather. Use only Circuit set B (set A was used for the review to highlight some common mistakes - the chap who drew them got a bit enthusiastic about the O.R. concept). I'll take a look - the issue remains short circuit protection though. There's no heating problem with the NTE11 in normal operation - the issue only comes about if there's a fault. Quote: If you are worried about current drain, the usual trick is put in a decent capacitor in series with a very low value resistor to give the current to move the solenoid, with a higher value resistor in parallel, to hold the coil in. No, no problem there at all actually..... the current draw is not significant, other than its effect on battery life. Quote: What batteries are you using? The only ones we found stand fast pressurisation and depressurisation are Lithium Ion Gel rechargeables. There are some quite nasty reports available for some other batteries in common use in rebreathers, one particular report by the US Navy comes to mind. Li Ion gel gives you tons of power. Charging circuit is in the above mentioned circuits also. Cheers Alex The batteries are going in a light can along with the main processor board and are both at 1 ATA and completely isolated from the breathing loop. I'm aware of some of the problems that can arise with batteries in a breathing loop or which have potential access to it..... LiIon gives you tons of power but also comes with some serious management issues and risks, as they are very ill-behaved if abused. I don't need that sort of power; NiMH works just fine. I'm getting plenty of runtime off a pack of 5 AA NiMH rechargeables - with the backlight off and no injection it approaches 100 hours before the low battery warning level is reached. With a "realistic" operational profile I'm seeing roughly 20 hours continuous use - vastly more than you can put on the unit during a diving day.
(Offline)

7th January 2006, 02:23	#26 (permalink)
WaTeRwEaSeL New Member Current Rebreather/s: Not Bought Yet Other Rebreather/s: Not Bought Yet Join Date: Dec 2005 Posts: 9	Re: Handset CAD anyone? :) Genesis, The handset looks great. Just a quick question on how you imported your CAD files onto the board. Did you transfer them to JPEG? I have a few CAD files I would like to post and I was just wondering. (Im using AutoCAD) Thanks Justin
(Offline)

7th January 2006, 02:27	#27 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,453	Re: Handset CAD anyone? :) I "printed" them to postscript (to a file) and then used Photoshop to convert to JPEG. I can also do a direct JPEG export from my CAD software if I want; it depends on what sort of detail you're after - the postscript conversion produces a much finer-grained image since its intended for printing at 300-600 dpi. If your software doesn't have a direct export the printing trick works with basically any package - just set up a "printer" that's a postscript to file output (use any common printer postscript driver under windows) and you're all set.
(Offline)

25th January 2006, 05:49	#28 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,453	Re: Handset CAD anyone? :) I've decided to pull the posts from the lid (it makes it MUCH easier to machine and the base material is much cheaper) as I found stand-offs that will work perfectly. Thus, the LCD will be mounted using nylon standoffs and machine screws to the lid, meaning the lid requires only the "lip" be machined and the holes drilled. I'll post an updated exploded view (I've also got the holes located for the magnetic switches as I've done a lot of testing now with both hall effect sensors and reed switches) soon - I've been tied up with boat work the last couple of weeks..... I'm waiting for some info from a couple of connector manufacturers before deciding whether to do the "entry hole" as a cable gland (easy and cheap) or use a connector. The connector (e.g. a fischer or one of the other competitors) has certain advantages, the most important being that a cable breach cannot propagate into the handset since the cable is "interrupted" at the connector..... __________________ "A venturesome minority will always be eager to get off on their own, and no obstacles should be placed in their path; let them take risks for Godsake, let them get lost, sunburnt, stranded, drowned, eaten by bears, buried alive under avalanches - that is the right and privilege of any free American." http://www.denninger.net http://www.diversunion.org/liability.htm - Fix the Diving Cert racket
(Offline)

25th January 2006, 14:10	#29 (permalink)
Scuba_Vixen e/mCCR Dolphin Pilot Current Rebreather/s: Other CCR Home Build Other Rebreather/s: Other CCR Home Build Join Date: May 2005 Location: St. Croix USVI Posts: 557	Re: Handset CAD anyone? :) Quote: (Originally Posted by Genesis) ~snip~ I'm waiting for some info from a couple of connector manufacturers before deciding whether to do the "entry hole" as a cable gland (easy and cheap) or use a connector. The connector (e.g. a fischer or one of the other competitors) has certain advantages, the most important being that a cable breach cannot propagate into the handset since the cable is "interrupted" at the connector..... I went with connectors for mine, for exactly the cable-breach/snag reason. Any problem with a cable and the controller remains unaffected. Cables are a lot cheaper to replace than a controller. As I posted previously, I liked the ones from Woodhead Connectivity, made by Brad Harrison. These are the same ones that Gene uses for the HS PO2 monitor. Well imagine my surprise to find what appeared to be the same ones in McMaster Carr, though with no mfg. mentioned. (I had been having a hard time sourcing them in small quantity) So I ordered a couple, and they are indeed the right ones, complete with the Woodhead identifcation and part numbers. I used the nano-change size for the HUD display, and the micro-change size for the cell holder. Would have liked the nanos for both, but had to work with a 20mm thread on the cell holder. Page 721 on McMaster site to see them I'll post a few pics when I get home from work and have time. Darlene
(Offline)

25th January 2006, 15:07	#30 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,453	Re: Handset CAD anyone? :) Cool - thanks..... will look into those.... __________________ "A venturesome minority will always be eager to get off on their own, and no obstacles should be placed in their path; let them take risks for Godsake, let them get lost, sunburnt, stranded, drowned, eaten by bears, buried alive under avalanches - that is the right and privilege of any free American." http://www.denninger.net http://www.diversunion.org/liability.htm - Fix the Diving Cert racket
(Offline)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode