SDRAM board
Moderators: Mug UK, Zorro 2, spiny, Greenious, Sorgelig, Moderator Team
Re: SDRAM board
I would like to buy the sdram board, but I can't find anybody selling them. I'm also looking at making them, what revision?
https://github.com/MiSTer-devel/Main_Mi ... mbly-(DIY) is this the latest?
https://github.com/MiSTer-devel/Main_Mi ... mbly-(DIY) is this the latest?
Re: SDRAM board
yesrobfish wrote: https://github.com/MiSTer-devel/Main_Mi ... mbly-(DIY) is this the latest?
-
- Obsessive compulsive Atari behavior
- Posts: 143
- Joined: Sat Oct 20, 2018 3:20 pm
Re: SDRAM board
Whats the difference between the SDRAM UNI and XS version please?
Re: SDRAM board
Mainly the height and orientation over the different revisions. the XS variant is supposedly a little faster operating.BlockABoots wrote:Whats the difference between the SDRAM UNI and XS version please?
-
- Obsessive compulsive Atari behavior
- Posts: 143
- Joined: Sat Oct 20, 2018 3:20 pm
Re: SDRAM board
thanks
Re: SDRAM board
I took a dive and ordered enough components to make 10 "Extra slim SDRAM Board v1.1" boards except I need 10 of these but they are out of stock:
2x20 (40 Pos.) Female Header, Double Row, Right Angle,
https://www.aliexpress.com/item/20Pcs-P ... 16130.html
Sorry if its been asked but is there another vendor I can order 10 of those from delivered to the USA? I'm looking on ebay and amazon but didn't know if anyone had a link to a known good version that would work.
2x20 (40 Pos.) Female Header, Double Row, Right Angle,
https://www.aliexpress.com/item/20Pcs-P ... 16130.html
Sorry if its been asked but is there another vendor I can order 10 of those from delivered to the USA? I'm looking on ebay and amazon but didn't know if anyone had a link to a known good version that would work.
Re: SDRAM board
Answering my own question but these look the same so I ordered them. Plus side its $1 cheaper and no shipping.HugoCraft wrote:I took a dive and ordered enough components to make 10 "Extra slim SDRAM Board v1.1" boards except I need 10 of these but they are out of stock:
2x20 (40 Pos.) Female Header, Double Row, Right Angle,
https://www.aliexpress.com/item/20Pcs-P ... 16130.html
Sorry if its been asked but is there another vendor I can order 10 of those from delivered to the USA? I'm looking on ebay and amazon but didn't know if anyone had a link to a known good version that would work.

https://www.ebay.com/itm/172272133044
Re: SDRAM board
I just assembled my first two ram boards, and wanted to share some feedback.
Firstly, thank you Sorgelig, I see you are doing an incredible amount of work on software/hardware. The open-source nature of this, right down to boards and SMD parts, as well as good write-ups have enabled me to put together a MiSTer even when all sellers are out of stock of parts. That's great!
I would say to improve docs/help more people, these doc changes could be made:
* State on DIY page that cores only require stability at 130MHz, but that as a guide 150MHz should generally be achievable with the recommended parts, while errors at 160+MHz does not indicate a problem. I know this information can be determined by reading around, but to have all in one place can give confidence/guidance quickly to first-time builders like me.
* As noted here, some of the KEMET capacitors are nearly impossible to find, and you can substitute for them, but the wiki lists "low ESR" type, and those seem to be less common (or less commonly indicated). Maybe a footnote about part substition might help. For example, is low ESR important for ram board? I am not an electrical engineer, I still don't know!
* A tip that usually these small SMD caps don't have polarity and can be soldered either way, same with clock crystal, would be nice.
* A few pictures of assembled board layout would be nice - I strained to check my IC was in right orientation since it was unmarked and I wasn't familiar with the text orientation rule.
* Some parts, like headers with long pins, I had surprising amount of trouble finding from a vendors who would ship quickly. I did find some, is there a place to submit links to good part substitions? At least, I think it would be helpful to make sure there is at least one link to parts that will ship faster than Ali, if known.
At least, keep these things in mind when the page is next updated. Since boards are in short supply tips to help self-assemblers who might not be experts are useful.
Firstly, thank you Sorgelig, I see you are doing an incredible amount of work on software/hardware. The open-source nature of this, right down to boards and SMD parts, as well as good write-ups have enabled me to put together a MiSTer even when all sellers are out of stock of parts. That's great!
I would say to improve docs/help more people, these doc changes could be made:
* State on DIY page that cores only require stability at 130MHz, but that as a guide 150MHz should generally be achievable with the recommended parts, while errors at 160+MHz does not indicate a problem. I know this information can be determined by reading around, but to have all in one place can give confidence/guidance quickly to first-time builders like me.
* As noted here, some of the KEMET capacitors are nearly impossible to find, and you can substitute for them, but the wiki lists "low ESR" type, and those seem to be less common (or less commonly indicated). Maybe a footnote about part substition might help. For example, is low ESR important for ram board? I am not an electrical engineer, I still don't know!

* A tip that usually these small SMD caps don't have polarity and can be soldered either way, same with clock crystal, would be nice.
* A few pictures of assembled board layout would be nice - I strained to check my IC was in right orientation since it was unmarked and I wasn't familiar with the text orientation rule.
* Some parts, like headers with long pins, I had surprising amount of trouble finding from a vendors who would ship quickly. I did find some, is there a place to submit links to good part substitions? At least, I think it would be helpful to make sure there is at least one link to parts that will ship faster than Ali, if known.
At least, keep these things in mind when the page is next updated. Since boards are in short supply tips to help self-assemblers who might not be experts are useful.
Re: SDRAM board
And resistors as well.diggie wrote:A tip that usually these small SMD caps don't have polarity and can be soldered either way, same with clock crystal, would be nice.
Every IC has mark of pin 1. Look more precisely and you will find the mark.diggie wrote: I strained to check my IC was in right orientation since it was unmarked and I wasn't familiar with the text orientation rule.
It's all on Wiki. When Wiki has more than one page there will be always some people reading it in different order and won't be satisfied. Amount of opinions equals amount of people.diggie wrote:* State on DIY page that cores only require stability at 130MHz, but that as a guide 150MHz should generally be achievable with the recommended parts, while errors at 160+MHz does not indicate a problem. I know this information can be determined by reading around, but to have all in one place can give confidence/guidance quickly to first-time builders like me.
Most components can be replaced, especially passives. It's community based Wiki and there is no people who is in charge of keeping it up to date. So, i'm not sure to who you appeal. I'm personally not going to watch it and keep it updated.diggie wrote:At least, keep these things in mind when the page is next updated. Since boards are in short supply tips to help self-assemblers who might not be experts are useful.
BOM was updated by people who decided to do it using links active at that time.
Generally speaking, i release only schematics and PCB files. BOM is out of my business/interest - i source all passives in local store. Some one can make it later but it won't exactly will happen.
Re: SDRAM board
I've made up a batch of SDRAM XS boards and have a few leftover, £15 each + whatever postage you would like
Thanks to all those contributing to this project =]

Thanks to all those contributing to this project =]

Please do not take it personal, my lawyer has policies
Re: SDRAM board
Hey guys,
I'm working on building some XS boards for those that need. I went with some Winbond chips and 1uf caps (Murata GRM21BR71C105KA01L) after having better luck versus .1uf (after someone said they had better luck). Im getting very different results with the boards - some pass 160, others only 140 (fortunately none are going lower). I figure with all the precautions im taking (ESD protection, iron tempurature, sourcing the ICs and caps from reputable vendors) I should be able to acheive 150/160 on all. Am I striving for too much "perfection" or am I missing some small details or tricks?
- What caps are others using for Winbond chips? (I see someone mentioned different cap values, but didnt mention which positions these caps belong on the board)
- Are people having better luck with Alliance chips versus Winbond?
I'm working on building some XS boards for those that need. I went with some Winbond chips and 1uf caps (Murata GRM21BR71C105KA01L) after having better luck versus .1uf (after someone said they had better luck). Im getting very different results with the boards - some pass 160, others only 140 (fortunately none are going lower). I figure with all the precautions im taking (ESD protection, iron tempurature, sourcing the ICs and caps from reputable vendors) I should be able to acheive 150/160 on all. Am I striving for too much "perfection" or am I missing some small details or tricks?
- What caps are others using for Winbond chips? (I see someone mentioned different cap values, but didnt mention which positions these caps belong on the board)
- Are people having better luck with Alliance chips versus Winbond?
Re: SDRAM board
The Alliance are quite a little bit 'better' according to noise it seems.
All of my Alliance have worked fine with .1u while for the Winbond I really would recomment 1u (I've had some boards where I have to exchange one of the 1u even to 10u to get them at least to 160MHz).
Yes the Alliance seems to be a little bit better but the price is 2.5 times high.
If your boards are in the 140MHz range - check your soldering .... might be that one pin just isn't soldered good enough.
All of my Alliance have worked fine with .1u while for the Winbond I really would recomment 1u (I've had some boards where I have to exchange one of the 1u even to 10u to get them at least to 160MHz).
Yes the Alliance seems to be a little bit better but the price is 2.5 times high.
If your boards are in the 140MHz range - check your soldering .... might be that one pin just isn't soldered good enough.
Re: SDRAM board
Interesting. I have been using all 1uf caps for the winbond. What cap do you typically change out with a 10uf to get 160?
So far all of my boards are only achieving 140mhz for a long period of time. They last maybe 10min at 150mhz. Even ones that previously tested at 160mhz. Im wondering if my DE10 is somehow acting up - but I have no known way to validate that. Or my environment is maybe adding noise (though I have this connected through an APC power filter). I wish the memtest gave more info about the errors it has encountered (like the x86 memtest).
I think my soldering is fine as I have even tried to reflow the chips with a little more solder and flux. Same result. I have done Xbox ram upgrades so I'm used to SMD stuffs like TSOP and QFP packages.
I will try a few Alliance chips to see if I can get any different result. I source them from a USA vendor so the Winbonds are only slightly cheaper than the Alliance chips. I may let these Winbond boards go a little cheaper if I can only get them to 140mhz.
Thanks for all the info guys. I'm really trying to avoid being the annoying n00b on this forum
So far all of my boards are only achieving 140mhz for a long period of time. They last maybe 10min at 150mhz. Even ones that previously tested at 160mhz. Im wondering if my DE10 is somehow acting up - but I have no known way to validate that. Or my environment is maybe adding noise (though I have this connected through an APC power filter). I wish the memtest gave more info about the errors it has encountered (like the x86 memtest).
I think my soldering is fine as I have even tried to reflow the chips with a little more solder and flux. Same result. I have done Xbox ram upgrades so I'm used to SMD stuffs like TSOP and QFP packages.
I will try a few Alliance chips to see if I can get any different result. I source them from a USA vendor so the Winbonds are only slightly cheaper than the Alliance chips. I may let these Winbond boards go a little cheaper if I can only get them to 140mhz.
Thanks for all the info guys. I'm really trying to avoid being the annoying n00b on this forum

Re: SDRAM board
The DE-10 does play a big part. I have one that won't pass any SDRAM board memory test at even 140MHz, and I have one DE-10 that passes at 167MHz for days with any of the modules I make.
You do not have the required permissions to view the files attached to this post.
I am the flux ninja
Re: SDRAM board
That is so strange. I'm amazed at the wide tolerance between DE10 boards. Does anyone have an idea of why that is?
I have ordered a couple Alliance chips from digikey, so we will see if they make any difference on my particular DE10.
In the mean time I will finish assembling my IO boards lol.
I have ordered a couple Alliance chips from digikey, so we will see if they make any difference on my particular DE10.
In the mean time I will finish assembling my IO boards lol.
Re: SDRAM board
I just saw a image of a sdram board with much bigger caps then what was listed on the BOM. I also have to mention that I have not successfully soldered a working sdram board. Does anybody have a part number?
Re: SDRAM board
Do you have a link to this image? The BOM calls only for .1uf, but for the Winbond chips I have had better~ish luck with 1uf.robfish wrote:I just saw a image of a sdram board with much bigger caps then what was listed on the BOM.
I've got some slightly diffrrent caps on order that I will try with my boards that are only getting 140mhz. (that said I'm not an electrical engineer, so its for S&Gs mostly).
Re: SDRAM board
Thank you,
just ordered 1uf caps
just ordered 1uf caps
Re: SDRAM board
One thing that I have seen with the Winbond chips is that they are very sensitive to capacitance created by solder flux residue between pins - much more sensitive than the Alliance brand. Use a LOT of solder flux remover and a fine bristled toothbrush (or similar brush) to remove any debris between pins. Even though I use a stencil, place the RAM with a microscope, and use a reflow oven, I still get a small amount of residue from the no-clean solder paste that I use. Whenever I have a Winbond RAM board that won't pass 167MHz I just re-clean it and then it passes.
I am the flux ninja
Re: SDRAM board
Excellent info. I'll try that. Thanks!JimDrew wrote:One thing that I have seen with the Winbond chips is that they are very sensitive to capacitance created by solder flux residue between pins - much more sensitive than the Alliance brand. Use a LOT of solder flux remover and a fine bristled toothbrush (or similar brush) to remove any debris between pins.
Re: SDRAM board
While I am waiting for flux remover (stuff has to ship ground), I tried one board with an Alliance chip. Strangely I get very similar results. Still only stable at 140mhz for 60min, but the error count seems to jump by large values each time (unlike the Winbonds, which had a very clear increase by "1" error periodically). This supports my concern the variance on my particular DE10 only supports 140ish mhz (as JimDrew alluded to).
I will still try flux remover as a test, and I may try to test my DE10 again in a different environment (maybe there is a lot of noise in my small basement office).
I will still try flux remover as a test, and I may try to test my DE10 again in a different environment (maybe there is a lot of noise in my small basement office).
-
- Obsessive compulsive Atari behavior
- Posts: 133
- Joined: Wed Aug 03, 2005 11:45 am
- Location: Ohio, USA
Re: SDRAM board
A good flux cleaning job does seem to be correlated with better performance, at least on the Winbond ones I've been building.
Almost all hit 167 on my DE10.
Almost all hit 167 on my DE10.
US based seller MiSTer Expansion Boards and Atari items
https://www.legacypixels.com/
https://www.legacypixels.com/
Re: SDRAM board
That's good to hear. There is no info in the datasheet pertaining to pin impedance or the effects of stray capacitance, but its pretty apparent that the Winbond chips are very touchy. The Alliance chips don't seem to care too much about solder flux residue.
I am the flux ninja
Re: SDRAM board
How did those replacement parts work out? I am also looking for the same part.HugoCraft wrote:
Answering my own question but these look the same so I ordered them. Plus side its $1 cheaper and no shipping.
https://www.ebay.com/itm/172272133044
Cheers!