Jump to content
IGNORED

IDE card tentative commitment page


marc.hull

Recommended Posts

I've already set up a spot on near the connectors to connect a 4-pin power plug of the 3.5" floppy-drive style, Swim. I also put two connectors so that it would be easy to attach a CF adapter at the back and still have a place to attach an IDE cable for a second drive. Note they are on the same IDE bus, so the maximum number of devices is still two for the card.

 

The area for the clock is still in flux, so any useful ideas here can still be implemented (with a greater or lesser degree of ease, depending on the chip). Many thanks for the input--I definitely can use the help! :)

Okay, I'm having a bad day. I replied to this already today but it seems to have gotten lost and never showed up. I'll try again.

 

One open/closed jumper that would be nice, Ksarul, would be for the bq4852 and bq4842 clock solutions. The signal at pin#21 of the bq4852 and pin#19 of the bq4842 are the signal to start the clock/battery of those chips. For safety sake, everyone has been breaking those pins off the chip after initial startup. If a jumper could open that line we wouldn't have to ruin the chips by breaking off the pin.

 

post-36289-0-55031400-1396978315_thumb.jpg

Link to comment
Share on other sites

I agree that the clock function is nice so please don't consider this a rail against it.

 

Problem is that only one of those chips is/was available in any (small) quantity (can't remember which but it is buried in the thread). Perhaps a good future proofing would be to include a space on board for a coin cell as IIRC the placement of the 256512 is all ready there ? Clock or no I'll still participate but I also vote for the option to utilize memory and battery backup for the day it is needed.

 

Another change that I personally would like to see is the elimination of the rotary CRU selector in favor of a DIP flavor. Since this is through hole that may all ready be done ?

 

Thanks for taking the time guys ;-).....

Link to comment
Share on other sites

I'll do the dip selector for sure, Marc--and for the clock chips, right now I have it set up so that you put whatever chip you want in the socket and jumper each pin across to the appropriate signal on the board. It would require up to 40 jumpers to connect the clock--but it is also pretty much future-proof, as any DIP clock chip can be put in there. . .I'll look at the one you suggested too, Swim. Again, many thanks for the input.

 

Gazoo, I'm not sure I'll have room to put a CF socket on there--but I'll look at the idea. Doing so might enlarge the rear extension of the card too much for stability Personally, I'd also avoid moving power to pin 20 on the connector, but I'll think about doing that too--or at least setting up a jumper for it.

  • Like 1
Link to comment
Share on other sites

I'll do the dip selector for sure, Marc--and for the clock chips, right now I have it set up so that you put whatever chip you want in the socket and jumper each pin across to the appropriate signal on the board. It would require up to 40 jumpers to connect the clock--but it is also pretty much future-proof, as any DIP clock chip can be put in there. . .I'll look at the one you suggested too, Swim. Again, many thanks for the input.

 

Gazoo, I'm not sure I'll have room to put a CF socket on there--but I'll look at the idea. Doing so might enlarge the rear extension of the card too much for stability Personally, I'd also avoid moving power to pin 20 on the connector, but I'll think about doing that too--or at least setting up a jumper for it.

I'll. Relay a recent experience. A while back I built an XT IDE card with an integrated CF adapter. It was surface mount and near impossible to do. According to the BOM and instructions this is the only way it's available.

Link to comment
Share on other sites

I'll. Relay a recent experience. A while back I built an XT IDE card with an integrated CF adapter. It was surface mount and near impossible to do. According to the BOM and instructions this is the only way it's available.

 

I second that. I built an IDE board for an ELF 2000 computer and it was a minor miracle that the IDE connector actually worked given how much trouble I had soldering it on the board. Definitely not worth it...

Link to comment
Share on other sites

One open/closed jumper that would be nice, Ksarul, would be for the bq4852 and bq4842 clock solutions. The signal at pin#21 of the bq4852 and pin#19 of the bq4842 are the signal to start the clock/battery of those chips. For safety sake, everyone has been breaking those pins off the chip after initial startup. If a jumper could open that line we wouldn't have to ruin the chips by breaking off the pin.

 

I'm intrigued! On the bq4852, pin 21 is D5 (according to the datasheet that I found). The battery is disconnected until the first application of Vcc on pin 36. What are people breaking pins off for after initial startup?

 

Stuart.

Link to comment
Share on other sites

  • 1 month later...

I've already set up a spot on near the connectors to connect a 4-pin power plug of the 3.5" floppy-drive style, Swim. I also put two connectors so that it would be easy to attach a CF adapter at the back and still have a place to attach an IDE cable for a second drive. Note they are on the same IDE bus, so the maximum number of devices is still two for the card.

 

The area for the clock is still in flux, so any useful ideas here can still be implemented (with a greater or lesser degree of ease, depending on the chip). Many thanks for the input--I definitely can use the help! :)

 

This week/weekend I've been using my IDE card more extensively. The biggest "gripe" I have with the card is how easily the DSR can be trashed. Once trashed, if the powerup routine finds the card, your system is hung out to dry until you reload the DSR, which often entails "hiding" the card until you can load the DSR loader.

 

Seems it would be worthwhile to split the DSR and RAM into two separate chips. Maybe use a flash EEPROM like some of the later SNUG cards. Either that, or come up with a hardware decode of the single RAM chip to enforce a write-protect on the DSR code banks while leaving the buffers read/write.

 

IMHO, the RAM-based DSR is a great idea but leaves the card very susceptible to failure and user frustration.

Edited by InsaneMultitasker
  • Like 1
Link to comment
Share on other sites

Maybe use a flash EEPROM like some of the later SNUG cards.

Yes, that works very well. Now all we need is someone who can write the code. ;)

 

Lucky for us IDE users, the CF doesn't get trashed when the DSR gets trashed.

 

At least that has never been the case for me.

Link to comment
Share on other sites

Yes, that works very well. Now all we need is someone who can write the code. ;)

 

Lucky for us IDE users, the CF doesn't get trashed when the DSR gets trashed.

 

At least that has never been the case for me.

Good point. I have not lost data, to my knowledge, resulting from the DSR getting trashed. The worst that happens to me is DM2K periodically trashes a file during recursive folder copying, resulting in an unusable target folder.

 

As for writing the code, I guess we need Fred's input before any changes could be entertained. If he writes to the DSR banks that would complicate things further.

 

Ah well, back to the real iron for some rare coding...

Link to comment
Share on other sites

  • 2 months later...
  • 2 years later...

You only have two choice with this one at the moment: take the Gerbers from Thierry's site and have a small number of them made (I've gotten good results using Sitopway off of eBay) or wait until I have time to finish doing a through-hole version of the layout. . .but that latter part may be a while.

  • Like 1
Link to comment
Share on other sites

I'm on board for 5 cards with parts.

 

(Side note; I just had to get two rails of the MBP clock chip because that's how many chips it took to get a decent price. Two rails of these chips will last the community into the next century but TIers still don't like to pay the high per chip price from the chip houses or ebay. We should probably just put these cards up as a kit so parts can be bought in quantity and then the per chip price won't prevent people from buying the card itself. Just my thought.)

  • Like 2
Link to comment
Share on other sites

I may just try to do a short run of these boards now (assuming the Gerbers from Thierry are in 274X format and not 274D format). I'll have to look into getting a parts list together and see if I can build a parts kit. I know there are a few HTF parts in the list, but I actually have a few of those that I bought just because they were available. Note that this wouldn't be a run of through-hole boards--they'd be Thierry's original SMT design. I am just not fare enough along in the conversion to through-hole to give any kind of a time-frame on that one (the change in chip footprint makes it an interesting dance to get all of the traces where I need them to be--and I do all of that work manually, not with routing software, as it is a lot more fun doing the traces manually).

  • Like 1
Link to comment
Share on other sites

Is the chip on the MBP one of the options for use with the IDE board? If it is, that will quickly suck up your stock of clock chips, Swim!

 

Actually the NATIONAL MM58167AN Real Time Clock 24-Pin Dip is not an IDE clock chip and would require it's own little circuit to dial it in so it keeps accurate time.

Link to comment
Share on other sites

I may just try to do a short run of these boards now (assuming the Gerbers from Thierry are in 274X format and not 274D format). I'll have to look into getting a parts list together and see if I can build a parts kit. I know there are a few HTF parts in the list, but I actually have a few of those that I bought just because they were available. Note that this wouldn't be a run of through-hole boards--they'd be Thierry's original SMT design. I am just not fare enough along in the conversion to through-hole to give any kind of a time-frame on that one (the change in chip footprint makes it an interesting dance to get all of the traces where I need them to be--and I do all of that work manually, not with routing software, as it is a lot more fun doing the traces manually).

 

The clock chips for the original cards are getting a little bit hard to find in the chip houses so it's best to just leave the board as is and do a run so any one of the three options can be used when found. As we already discussed earlier in this thread the RTC65271 unit may be the one to shoot for in a parts kit.

Link to comment
Share on other sites

Doing some due diligence, of the four clock chips out there, the two that require an external RAM chip (the RTC65271 and the BQ4847) are available in the $20 per chip range. As Swim noted, the RTC65271 is a bit of a better option, as the batteries are replaceable. The BQ4842 (128K RAM) is pretty much unobtainium now. A lot of places have the BQ4852, but most of them want $220+ per chip. I have one source where I have an account that sells them for about $70 each though--and as that source is an authorized distributor, the chips I'll get from them are genuine. The same source also has the BQ4847, in the already mentioned price range. The SRAM chips were about $20 each the last time I bought some of the right type--and I got those for a steep discount. Of the two alternatives, the one from ST Micro shows as available at three parts houses--without a posted price, so I'd have to submit an RFQ to find out their strike price. Nobody reliable has the Toshiba chip, although a few parts houses in Hong Kong and China claim to have them. That puts the cost of the RTC+SRAM in striking distance of the Battery-backed RTC with onboard SRAM--and the latter is a guaranteed new part, as my source is authorized to manufacture new ones by the OEM. That would put the parts kit into the $150 to $200 range and require the buyers to build the boards themselves. I did find some of a slightly different (compatible) Toshiba chip just now though--and the seller has those for $5 each, so I may try some of them with the RTC65271, as that is definitely a cheaper option than all of the others I've seen. I'll have to look and see what remaining parts happen to be pricey to see what the total cost is, but these will at least lower it below the $150 range, making that the worst-case cost for a kit.

  • Like 1
Link to comment
Share on other sites

That's what I'm working on. The RTC isn't a problem, I can get those in quantities up to 1,000 or so, but the price break isn't much between 25 and 1,000 chips. At 25 chips, it goes below $20, at 100 it drops to $19 and at 1,000 it drops to $18 or so. Not much of a break at all. . .

 

Right now I'm working on the SRAMs, as those are the long-pole item. The few sources I've found have them in limited numbers for acceptable prices, so I'm working on hoovering up as many as I can get to see what my average price per chip will be.

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...